Music, Text, and Culture in Ancient Greece

OUP CORRECTED PROOF – FINAL, 30/1/2018, SPi
MUSIC, TEXT, AND CULTURE

I N A N C I E N T GR E E C E
Music, Text, and

Culture in Ancient
Greece
Edited by
TOM PHILLIPS
and
A R M A N D D ’A N GO U R
1
3
Great Clarendon Street, Oxford, OX2 6DP,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© Oxford University Press 2018
The moral rights of the authors have been asserted
First Edition published in 2018
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Published in the United States of America by Oxford University Press
198 Madison Avenue, New York, NY 10016, United States of America
British Library Cataloguing in Publication Data
Data available
Library of Congress Control Number: 2017955710
ISBN 978–0–19–879446–2
Printed and bound by
CPI Group (UK) Ltd, Croydon, CR0 4YY
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
Preface
This volume originates in a workshop that took place in June 2013,

and a conference that took place a year later. We would like to express
our gratitude to the John Fell Fund for making these events possible,
and to Jesus College, Oxford for providing an atmosphere congenial
to discussion. The spirit of lively debate and co-operative engagement
that pervaded those meetings has carried through the process of
preparing the volume for publication, and we are very grateful to
the contributors for their patience and efficiency. Georgina Leighton
and Charlotte Loveridge steered the volume through the press with
attentiveness and skill, and the final product was considerably
improved by the editorial interventions of Tim Beck and Albert
Stewart. Numerous other scholarly conversations have informed
the volume, but we are particularly indebted to Emily Dreyfus and
Pauline LeVen for their comments on the introduction, and to the
readers of the press for their suggestions about the shape and sub-
stance of the book as a whole.
A.J.D.
T.R.P.
Oxford
October 2016
Contents
List of Figures ix
List of Abbreviations xi
List of Contributors xiii
Introduction: Music, Text, and Culture 1

Tom Phillips
P A R T I . I N T E R P R E TA TI O N
1. Epicentric Tonality and the Greek Lyric Tradition 17
John C. Franklin
2. The Musical Setting of Ancient Greek Texts 47
Armand D’Angour
3. Words and the Musician: Pindar’s Dactylo-Epitrites 73
Tom Phillips
4. Music in Euripides’ Medea 99
Oliver Thomas
5. Mesomedes’ Hymn to the Sun: The Precipitation
of Logos in the Melos 121
Stelios Psaroudakes
P ART II. THE ORY, R ECEP TION, CONTEXTS

6. Hearing the Syrinx in Euripidean Tragedy 139
Naomi Weiss
7. Lyric Atmospheres: Plato and Mimetic Evanescence 163
Anastasia-Erasmia Peponi
8. Aristotle on Music for Leisure 183
Pierre Destrée
9. Sounds You Cannot Hear: Cicero and the Tradition
of Sublime Criticism 203
James I. Porter
viii Contents
10. Disreputable Music: A Performance, a Defence, and their
Intertextual and Intermedial Resonances
(Plutarch Quaest. conv. 704c4–705b6) 233
Andrew Barker
Bibliography 257
Index 277
List of Figures
1.1. Oscilloscope readings of two tones at unison, the 3:2 ‘fifth’, the
4:3 ‘fourth’, and so on through the 9:8 wholetone (Gk τόνος). 21
1.2. Graphic representation of the Mesopotamian tuning cycle
UET VII 74. 23
1.3. Epicentric arrangement of traditional Greek heptachord,
according to Aristotle. 26
Drawn by Bo Lawergren and originally published in
B. Lawergren (1998) ‘Distinctions among Canaanite, Philistine,
and Israelite Lyres, and their Global Lyrical Contexts’, Bulletin
of the American Schools of Oriental Research 309, 41–68.
1.4. ‘Recomposition’ of Sappho 1, illustrating epicentric tonality
and accent-melody in conjunct heptachord E-F-G-A-Bb-c-d. 39
1.5. Didymus’ chromatic γένος, expressed in matrix showing ratios
between all string pairs, with decimal figures replaced by
resonant/epimoric ratios (3:2, 4:3, 5:4, etc.) where applicable. 43
5.1. Melodic diagram of Mesomedes’ Hymn to the Sun. 131
List of Abbreviations
CBS Catalogue of the Babylonian Section, University Museum,

Philadelphia.
CEG P. A. Hansen (ed.), Carmina Epigraphica Graeca (Berlin, 1983–9).
DAGM E. Pöhlmann and M. L. West (eds), Documents in Ancient Greek
Music (Oxford, 2001).
D-K H. Diels and W. Kranz (eds), Die Fragmente der Vorsokratiker,
I–III (Berlin, 1974).
Dr A. B. Drachmann (ed.), Scholia Vetera in Pindari Carmina, I–III
(Leipzig, 1903–27 [reprinted Stuttgart, 1997]).
FGrH F. Jacoby et al. (eds), Fragmente der Griechischen Historiker
(Leiden, 1923–).
K-A R. Kassel and C. Austin (eds), Poetae Comici Graeci (Berlin, 1983–).
LfgrE Lexikon des frühgriechischen Epos (Göttingen, 1979–).
LIMC Lexicon Iconographicum Mythologiae Classicae, I–VIII (Zurich/
Munich, 1981–99).
LSJ H. G. Liddell, R. Scott, H. S. Jones, and R. Mackenzie (eds), A
Greek–English Lexicon (ninth edn, Oxford, 1940).
MSG K. von Jan, ed. Musici scriptores Graeci: Aristoteles, Euclides,
Nicomachus, Bacchius, Gaudentius, Alypius (Leipzig, 1895).
PEG A Bernabé (ed.), Poetarum epicorum Graecorum testimonia et
fragmenta I (Leipzig, 1987).
PMG D. L. Page (ed.), Poetae Melici Graeci (Oxford, 1962).
PMGF D. L. Page and M. Davies (eds), Poetarum Melicorum Graecorum
Fragmenta (Oxford, 1991).
SEG Supplementum Epigraphicum Graecum (Amsterdam, 1923–).
S-M B. Snell and H. Maehler (eds), Pindari Carmina cum Fragmentis
(Lepizig, 1984, 1989).
TLG Thesaurus Linguae Graecae (University of California, 1985–).
TGrF R. Kannicht, S. Radt, and B. Snell (eds), Tragicorum Graecorum
Fragmenta, I– (Göttingen, 1971–2004).
UET Ur Excavation Texts (London, 1928–).
VAT Vorderasiatisches Museum, Berlin (Vorderasiatische Abteilung.
Tontafeln).
Wehrli F. von Wehrli (ed.), Die Schule des Aristoteles: Texte und
Kommentar I–X (Basel and Stuttgart, 1969).
List of Contributors
Andrew Barker is Emeritus Professor of Classics at the University of

Birmingham. He has published eight books and numerous articles on
ancient Greek music and musical theory, and is the Founding Editor
of the journal Greek and Roman Musical Studies. He was elected a
Fellow of the British Academy in 2005.
Armand D’Angour is Associate Professor in Classics at Oxford and
Fellow and Tutor of Jesus College. He is the author of The Greeks and
the New (Cambridge, 2011), and has published numerous articles on
ancient Greek music and poetry. His ongoing project aims to recon-
struct the sounds of ancient Greek music.
Pierre Destrée is a FNRS Research Professor at the University of
Louvain. Most recently he has co-edited The Blackwell Companion to
Ancient Aesthetics (2015), and Plato: Symposium—A Critical Guide
(Cambridge, 2017).
John C. Franklin is Associate Professor and Chair of Classics at the
University of Vermont. The cultural history of ancient music has
been central to his research, much of which has focused on the
interface between early Greece and the Near East (culminating
recently in Kinyras: The Divine Lyre, 2016).
Anastasia-Erasmia Peponi is Professor of Classics at Stanford
University. She writes on issues of aesthetic perception and judgement,
ancient and modern lyric poetry, Plato, dance, and the relationship
between the verbal and the visual. Among her publications are Fron-
tiers of Pleasure: Models of Aesthetic Response in Archaic and Classical
Greek Thought (Oxford, 2012) and (as editor) Performance and
Culture in Plato’s Laws (Cambridge, 2013).
Tom Phillips is Supernumerary Fellow in Classics at Merton College,
Oxford. He is the author of Pindar’s Library: Performance Poetry and
Material Texts (Oxford, 2016). His current research focuses on lyric
poetry, Hellenistic poetry, and ancient scholarship.
James I. Porter is Chancellor’s Professor of Rhetoric and Classics
at UC Berkeley. His teaching and research focuses on models
xiv List of Contributors

of aesthetic sensation, perception, and experience in ancient Greece
and Rome. His most recent book is The Sublime in Antiquity
(Cambridge, 2016).
Stelios Psaroudakes is Assistant Professor of Ancient Hellenic Music
in the Department of Music Studies of the National and Kapodistrian
University of Athens. He teaches and researches in the areas of
ancient music theory, notation, scores, and organology.
Oliver Thomas is an Assistant Professor in Classics at the University
of Nottingham. He is the author, with David Raeburn, of The Aga-
memnon of Aeschylus: A Commentary for Students (Oxford, 2011).
His research currently centres on Greek hymns, and he is completing
a commentary on the Homeric Hymn to Hermes for Cambridge
University Press.
Naomi Weiss is Assistant Professor of the Classics at Harvard
University. Her research focuses on ancient Greek performance culture,
especially theatre. She is the author of The Music of Tragedy: Perform-
ance and Imagination in Euripidean Theater (forthcoming) and is
currently co-editing a volume on archaic and classical lyric genres.
Introduction
Music, Text, and Culture
Tom Phillips
This volume addresses two issues central to the study of ancient

Greek performance culture. The first is the narrow but methodologic-
ally problematic question of the role played by music in performed
poetry. The second is broader: how did the ancients understand the
relationship between music, poetry, and performance, and how did
reflection on music relate to other areas of ancient intellectual life?
While the second has received considerable attention, the first, des-
pite its obvious importance, remains thinly formulated and little
understood. There are obvious and intransigent reasons, both insti-
tutional and evidential, for this interpretative blindspot. Perhaps
more clearly than any other area of scholarship the study of music
in ancient Greece exemplifies both the benefits and the problems of
increasing scholarly specialization. Recent years have seen scholars
produce works of great technical sophistication that have vastly
increased our understanding of ancient instruments, modal systems,
and musical scholarship,1 and yet the difficulty of the material these
works address has meant that music has generally remained some-
thing of a closed book to mainstream literary scholars. Formidable
evidential problems also beset those interested in the role music
played in performance culture, especially in the archaic and classical
periods. While inscriptions and papyri furnish considerable evidence
1
See e.g. West (1992b); Barker (2007); Hagel (2009); Creese (2010). For an
accessible overview of ancient musicology see Barker (2014).
2 Tom Phillips
for musical culture in the Hellenistic period and beyond,2 our evi-
dence for the music of Homer, the tragic and comic choruses, and the
choral genres of the sixth and fifth centuries, not to mention the
various forms of monody and popular songs, is for the most part
exiguous. As a result, it has become a scholarly topos to pair an
emphasis on music’s importance to performance poetry with
acknowledgement of our ignorance of its workings.3 On the other
hand, the social, ritual, and political contexts of poetic performance in
the classical period are better represented in our evidence, and great
advances have recently been made in studying performance ‘in the
round’, as a phenomenon that intersects with various other social
factors.4 The social environments in which music was composed,
performed, listened to, and debated are consequently much better
understood than they were a generation ago.5 One implication of this
volume, however, is that classical and archaic poetry, and the ancient
sources that bear on it, still have much to tell us about how music
worked in the archaic and classical periods, even though interpreting
the evidence requires a precarious mixture of imagination and
caution.
A central aim of the chapters collected here is to make connections
between musicological scholarship and the issues that have tradition-
ally concerned students of ancient literature. Especially pressing
in this respect is the need to develop a better understanding of
how music and texts combined in performance, and it is this to
which the essays in the first half of the volume are largely devoted.6
The ‘music’ of our title and the use of ‘music’ in these opening
remarks goes against the terminological grain of recent scholarship
2
On Hellenistic and later musical scholarship see e.g. Prauscello (2006).
3
See e.g. Gentili (1988) 31.
4
For classical choral lyric Calame (2001) is foundational; more recently see e.g.
Kowalzig (2007); Athanassaki and Bowie (2011); Fearn (2011); Kowalzig and Wilson
(2013). On tragic choruses see especially Gagné and Hopman (2013); for responses to
choral culture in Plato see Prauscello (2013a), and the essays in Peponi (2013a). For
musical culture more generally see Murray and Wilson (2004) and Yatromanolakis
(2011).
5
Cf. Csapo (2004) on the New Music; Power (2010) on the history of citharodia.
6
For other recent moves in this direction see e.g. Wilson (2005); Hall (2006)
288–320; Goldhill (2013); Phillips (2013); Gurd (2016): see further Porter (this
volume) 217 n. 36. Nooter (2012) analyses the shifts between spoken and sung
utterance in Sophocles; on the role of sound in poetry, looking especially at the
voice, see Butler (2015) 82–7.

on performance culture, which has tended to ground discussion of
ancient musico-cultural phenomena in terms such as μουσική or
μολπή.7 The manoeuvre of employing these ancient terms has the
value of highlighting the cultural differences between ancient and
modern music, especially as regards the ‘seamless complex of instru-
mental music, poetic word, and co-ordinated physical movement’8
that comprised μουσική in ancient performance culture. Our use of
‘music’ is not meant to signal a move away from considering musical
elements (the sounds of instruments, vocal melodies, etc.) as part
of performance totalities, but rather to recalibrate what analysis of
the musical element of μουσική entails. A persistent concern of the
first half of the volume is to isolate phenomena such as melody and
rhythm and consider them as distinctive evidential strands, enabling
them subsequently to be integrated into more holistic analyses of
performance. This more discrete treatment of μουσική enables the
distinctiveness of musical elements to emerge more clearly, but it
also promotes greater attention to how the verbal and the musical
interact.
This terminological choice and the interpretative manoeuvre that it
reflects are paralleled in contemporary reflection on the use of mul-
tiple media in artworks. Discussions of mediality—the properties and
qualities of the media in which artworks are constructed—has
addressed various artistic phenomena. A particular concern has
been the elucidation of relations between different media within a
single artwork. Another subject of focus has been the formal features
of one medium being transposed into or reworked in another.9
Verbal descriptions of music, and the adoption in novels of structur-
ing devices borrowed from music, have received considerable atten-
tion.10 In one sense, discussion of such interactions corresponds to
what classicists have long been doing in their treatment of phenomena
such as ecphrasis, but the specific terms in which these discussions are
framed also prompt us to reconsider the ways in which we think about
the relationships between media in artworks.
One of the animating principles of studies on mediality is the
‘recognition that the arts and media should not be studied in their
7
See e.g. Murray and Wilson (2004).
8
Murray and Wilson (2004) 1.
9
Defined as ‘transmediality’: for a discussion see Kattenbelt (2008) 23–4.
10
Cf. Wolf (2002) 23–5 for discussion and further references.
4 Tom Phillips
own historical developments and with their own rules and specifica-
tions, but rather in the broader context of their differences and
co-relations’.11 This mirrors to some extent the emphasis on a more
contextually grounded approach to archaic and classical Greek
poetry, which has emphasized that poetry is not a type of language
that belongs to its own separate realm, but needs to be seen in relation
to ritual practices, socio-political discourses, and the plastic arts. Yet
in emphasizing both ‘differences and co-relations’, the study of medi-
ality provides a (necessarily provisional) framework for addressing
both the combination and disaggregation of media within an art-
work.12 Works that include or are realized through multiple media
can be analysed in terms of the ‘fusion’ they create,13 but can also be
understood as creating medial interactions that change the signifi-
cance of individual media.14 Crucially, an intermedial approach
resists conceptualizing artworks in terms of a critical programme
in which the semantic is posited as the privileged model to which
musical, rhythmical, and other non-verbal aspects of an artwork
are subordinated, a tendency which has been common in critical
writing from antiquity to the present.15 In attempting to do justice
to the shifting multifariousness of the relationships between music
and text, semantics and prosody, the chapters in this volume share
the focus, if not always the language, of mediality studies. They
highlight instances of music and rhythm ‘imitating’ or reinforcing
semantic content,16 but they also explore moments at which texts’
11
Kattenbelt (2008) 20.
12
For a concise overview of different relations between media see Kattenbelt
(2008), and for more extended discussion Wolf (2002). Particularly germane to the
concerns of this volume are the comments of Wolf (2002) 17 on ‘intracompositional
intermediality’, which he defines as ‘a direct or indirect participation of more than one
medium of communication in the signification and/or semiotic structure of a work or
semiotic complex’. For further theorization on the relations between words and music
and further references see the essays collected in Bernhart (2012).
13
Wolf (2002) 20.
14
Cf. Kattenbelt (2008) 25, using ‘intermediality’ to refer to ‘co-relations between
different media that result in a redefinition of the media that are influencing each
other’. Such ‘redefinitions’ may be more or less radical depending on the nature of the
media involved: Kattenbelt compares Kandinsky’s Bühnenkompositionen, in which
the interplay of arts included within the whole was grounded in their perceived
independence, with Wagner’s conception of the Gesamtkunstwerk in which music
was given primacy.
15
See further Phillips (this volume) 74–9.
16
See e.g. the chapters by D’Angour and Psaroudakes in this volume.

auditory dimensions assert, or are understood as having, a distinctive
autonomy.17
When combined with traditional philological approaches, the con-
ceptual tools provided by theorization about mediality can enable a
more granular analysis of how words and music interact in poetic
texts, but they are also ripe for extension in the light of the medial
complexities of particular texts, as Andrew Barker’s discussion of
‘potential intermediality’ in Plutarch demonstrates.18 Equally, media
theory furnishes readers with a critical idiom that has greater typo-
logical than descriptive or interpretative resources.19 This is high-
lighted by the fact that the problem of finding adequate vocabulary
with which to register the emotive effects and interpretative chal-
lenges created by ‘multimedial’ artworks is especially pressing in the
case of ancient culture. In addition to the challenge of translating and
understanding ancient aesthetic terminology and intellectual frame-
works,20 we also have to grapple with our relative lack of relevant
acculturation. A contemporary reader of a poem entitled ‘Bob Dylan’s
Minnesota Harmonica Sound’ can be expected to have an intuitive
grasp of the subject derived from frequent encounters with the music,
mention of which will usher in, for such a reader, a set of memories
and personal associations which inflect their participation in the
poem’s imaginative world.21 When reading ancient texts in which
music plays a part, we are at best less familiar with the relevant
parallel phenomena, and at worst almost totally in the dark. The
sounds of the aulos, the shape of rhythmical phrases, and the tones of
voice employed in singing about intense emotional experiences
17
For the former see especially the chapters by Phillips and Thomas, for the latter
those by Peponi and Porter in this volume.
18
Pp. 249–55.
19
The discussion of Stimmung by Peponi (this volume, esp. 167–8, 172–4) exem-
plifies this. As conceptualized by Plato, the Stimmung of lyric poetry is multimedial in
the sense of being created by prosody and the dynamics of the individual voice. What
makes this Stimmung distinctive in its philosophical context, however, is not its
multimedial aspect per se but its production of ‘atmospheres’ that are both all-
pervasive and indeterminate, and Plato’s response to the critical problem these
atmospheres create.
20
See D’Angour (this volume) 48; Phillips (this volume) 74–81.
21
Lachlan Mackinnon, The Jupiter Collisions (Faber & Faber, 2003). The poem
itself, however, makes no explicit reference to its titular subject. This obliqueness
opens up various ways of construing the relationship between the ‘sound’ in the title
and the poem’s soundscape, dominated by ‘the wind that seemed to be always
blowing’ and the whistling of freight trains.
6 Tom Phillips
would all have contributed powerfully to the experience of listening
to a Euripidean choral ode. Ancient readers, even those of periods
fairly remote from Euripides’ own time, would have brought to a
reading of the same ode a much greater familiarity with these
phenomena than their modern counterparts,22 for whom they pose
ascendingly complex problems of reconstruction and interpretation.
We can never hope to hear ancient poetry as the ancients did. But in
scrutinizing the effects created by interactions between music and
text, the chapters in the first half of the volume aim to deepen our
understanding and imaginative engagement with this aspect of
poetry, as well as to connect these interactions with wider interpret-
ative questions.
Although they do not attempt to offer an overarching theory of the
relationship between words and sounds, these chapters share a focus
on the two-way nature of the relationship between music and texts.
The volume begins with an essay that examines this relationship
against a historical backdrop that reaches back to the third millen-
nium BC. John Franklin argues that the tuning system the Greeks
applied to the lyre derived ultimately from Assyrian and Babylonian
models. Franklin tracks this ‘epicentric’ model, which used the cen-
tral string (μέση) as the basis for determining the tunings of the
others through the archaic and classical periods, and finds signs of its
influence persisting into later antiquity.23 As well as being richly
suggestive for melodization in the archaic period, Franklin’s explor-
ation of the epicentric model demonstrates the effects of musical
practice on other cultural activities.24
Despite the variations in the contexts, genres, and communicative
strategies of the primary texts under discussion in the next four
chapters, a basic dynamic recurs: as Armand D’Angour and Stelios
Psaroudakes demonstrate in detail, music affects how individual
words and phrases are understood, while the verbal structures
inflect how their melodies, rhythms, and instrumental accompani-
ment affect listeners. As well as allowing a richer appreciation of the
formal structures of the texts under discussion, this approach
22
See also Barker (this volume) 251–2 on the challenges ancient readers faced in
conceptualizing descriptions of musical performances.
23
See e.g. pp. 44–5 on Plutarch’s treatment of the lyre’s boundary strings (lower,
middle, and upper) as Muses.
24
See esp. pp. 43–4.

enhances our understanding of musical affectivity. Rhythms and
musical phrases create and participate in small-scale textual
effects,25 but the cultural associations of individual instruments
and melodic forms enable musical accompaniment to affect inter-
pretation in a way that is all the more telling for being formally
distinct from the texts’ verbal content. Depending on the context
of the utterance, musical accompaniment can play a crucial role
in directing listeners’ interpretation of ethical, psychological, and
political issues.26
Another feature of these chapters informed by a focus on the
medial distinctiveness of performance poetry is their examination of
the boundaries between musical and non-musical elements. On the
reading of Pindar’s dactylo-epitrites advanced in Chapter 3, rhythm is
a powerfully affective element, but also ‘signifies’ in ways that are not
readily graspable in terms normally applied to verbal meaning.27
Similarly, Oliver Thomas’s interpretation of the Medea shows that
music could contribute to the characterization of choral utterance
and act as a vehicle for implicit authorial comment on the chorus’s
propositions and arguments.28 All four chapters envisage authors
composing musically complex works that presuppose a high level of
critical sophistication among listeners, and entail modes of under-
standing alert to both the distinctions and continuities between the
musical and verbal elements of performance poetry. As such, these
analyses illustrate what can be gained from the manoeuvre outlined
above of uncoupling μουσική and music. At the same time, their focus
on phenomena in which sound and sense combine also highlights
both the importance and the provisionality of the categorical distinc-
tions employed by ancient and modern criticism.
As well as offering new ways of thinking about the texts in ques-
tion, the approaches explored in these chapters contribute to larger
and enduringly difficult questions about music’s ontology and func-
tion. Thinkers ancient and modern have debated the role of sound in
25
On the role played by rhythm in framing responses to poetry see e.g. Thaut
(2005) 5; London (2002) 532–3; Prauscello (2013b) 258–9. For discussion of rhythm
in ancient Greek poetry see Edwards (2002), esp. 62–98; more recently, see the essays
in Celentano (2010).
26
Cf. Kramer (1990) 269–70 on music as ‘a participant in, not just a mirror of,
discursive and representational practices’, with the further discussion of Scher (1997)
14–15.
27 28
Phillips (this volume), esp. 89, 95–7. See esp. pp. 112–18.
8 Tom Phillips
producing semantic meaning,29 whether music can be understood to
function like a language,30 whether music can represent objects,
emotional states, or concepts,31 and whether the terms used to discuss
linguistic meaning are fundamentally inapposite to the description of
music. Although theoretical engagement with these debates is not a
principal aim of this volume, the interpretations of ancient poetry
elaborated throughout intersect in various ways with the wider ques-
tions with which these debates are concerned. Music’s representa-
tional capacities, for instance, are a recurring issue. The mimetic
aspect of music is a commonplace of the ancient sources, but a
particular feature of the readings collected here is that they illustrate
the multidimensional complexity of how mimesis actually worked
in performance. From one musical instrument masquerading as
another, to a melody suggesting the interpretative shortcomings of a
chorus’ viewpoint, or rhythmical sequences embodying ideas of
abstract order, the range of mimetic resources at work in these texts
resists unitary conceptualization.32 Study of texts from the kind of
multimedia perspective utilized in these analyses allows further meat
to be put on the bones of ancient sources that discuss musical
mimesis, but also enables engagement with modes of mimesis not
documented by ancient critics.
The second half of the volume addresses directly the issue of how
the ancients responded to and conceptualized music. The chapters
range widely in time and subject matter, discussing how concepts of
musicality and sound are manipulated, discussed, and challenged
across a variety of rhetorical, philosophical, and critical texts. Some
29
The distinction between sound generated by instruments or other types of
accompaniment and the sound of language itself is particularly important in this
respect: see especially Porter (this volume) for ancient literary critics’ interest in
this issue.
30
See e.g. Coker (1972) for the argument that music can convey concepts, a
position which can be seen as a reconfiguration of the ancient position which held
that music could represent character: see e.g. Destrée (this volume) 192 discussing
Arist. Pol. 8.1340a18–28.
31
See e.g. Davies (1994) 1–49, particularly 48. Koopman and Davies (2001) argue
for an ‘expanded’ notion of musical meaning; see Kivy (2007) for a statement of
music’s resistance to meaning in the linguistic sense. Some important objections to the
notion that music can have a representational function include Scruton (1976) and
Davies (1994). See Dreyfus (2013) 167–8 for a succinct assessment of the problems
with treating music as a representation of propositions that can be conveyed through
language.
32
See e.g. Ch. 6, pp. 112–15, and 95–7.

chapters focus on how ancient philosophers and literary critics
responded to the theoretical issues posed by the conjunctions of
music and text, while others examine aspects of music’s broader
cultural significance and the influence of musicology on disciplines
such as literary criticism. For all their variations in subject and
methodology, however, these chapters convey a vivid sense of the
intellectual challenges music posed to ancient audiences (and readers).
In antiquity as much as today, music was a feature of everyday life,
and yet (in some of its forms, at least) had a capacity for disturbing
and delighting listeners in ways that strained conceptual categories.
One function of this volume is to draw out continuities and contrasts
between ancient and modern theorizing about and experience of
music; identifying continuities in how music is conceived and debated
is a means (albeit provisional) of bridging the cultural divide that
separates us from antiquity. Equally, we can sometimes identify simi-
larities in melodic practice between ancient and modern texts which
suggest that the responses of (at least some) ancient listeners may not
have been very different from types of response with which we are
familiar.33 But in highlighting contextual and intellectual differences
between antiquity and today, the volume also enables a more focused
understanding of what made ancient musical culture distinctive.
The traffic between music and texts pivots around a fundamental
tension: despite its mimetic capacities, music’s significational
resources differ from those of a language. Although music can be
conventionally or critically framed as a representational mode, the
nature of its function is such that it always tends towards exceeding
determination in purely semantic terms.34 At the same time, however,
33
Caution is of course required in guarding against projecting our own assump-
tions about such practices back onto antiquity, but the formal parallelisms are often
suggestive: see D’Angour (this volume) 61, 71–2 for continuities between the gener-
ation of affect through melodic shape in ancient and modern music.
34
This has often been phrased in terms of the resistance of musical experience to
being captured in language. See e.g. Shepherd and Wicke (1997) 143, who argue that
‘music theory and music analysis are quite different and distinct in the character of
their thinking from the character of musical experience. They cannot “reach out” to
musical experience in any convincing or useful manner’; see also Jankelevitch (2003)
71–6, 102–3. Cf. however the remarks of Cook (2001) 189–90. Gurd (2016) addresses
this problem from a different angle by examining the tension between ‘noise’ and
‘sound’ in Greek poetics, the former being the irreducibly material element that
always threatens to overrun the constraints imposed by the formal systems which
produce the latter.
10 Tom Phillips
it is open to being marked by verbal meaning.35 This dialogue, in
which music both invites and resists accommodation to the seman-
tic,36 is in part a product of the conditions in which music is pro-
duced. When an aulos is played during a tragic choral ode, an
invitation to conjoin, or at least to relate, sound and semantics is
inscribed into the structure of the performance. A particularly com-
plex instance of this relationship is the subject of Naomi Weiss’s
essay on the syrinx in Euripides’ choral odes. Drawing on a distinc-
tion elaborated in contemporary sound studies between the actual
sounds made by instruments and how these sounds are imagined by
listeners, Weiss shows that the aulos could momentarily represent
the syrinx, allowing Euripides to mobilize the cultural associations
of both instruments, and the tensions between them, for a variety of
dramatic ends.
While Weiss focuses on how sound-making is represented, and on
the interpretative frames choral odes project for themselves by means
of these representations, other essays explore how interactions
between sound and meaning are informed by the differences between
their communicative modes. These interactions, with their potential
for turbulence and disruption as well as for the sensory reinforcement
of semantic content, are crucial to what the ancients termed music’s
‘ethical’ dimension. Music’s capacity to express emotions and to
engender emotional responses has been a central concern of ancient
and modern theory.37 Yet while modern musicologists have tended to
focus their attention on formal questions relating to what (if any)
emotional or conceptual content music can convey,38 the emphasis in
35
Cf. e.g. Kramer (1990); Cook (2001) 178–9.
36
Cook (2001) 187–8 suggests that ‘the instability of music as an agent of meaning’
derives from the fact that, although music can express nuance, ‘the necessary inter-
pretative decisions’ for judging such nuance are not given by the music itself. This
position can be compared with Peponi’s remarks on Plato (this volume, 174–8); for
further discussion see Phillips (this volume) 87–9.
37
Issues of emotiveness and ethical force are central to Plato’s remarks about
music: see e.g. Rep. 410a–412b, 423d–425a, Lg. 653c–660d.
38
Debate about musical emotiveness is closely connected to questions about its
representational functions. The modern bibliography on this subject is vast. Founda-
tional is Hanslick (1974) (first published 1854); his arguments have often been taken
to hold that music is a purely formal system without representational capacities,
but see the remarks of Cook (2001) 174–5. Cooke (1959) understands music as an
emotive language with certain basic constituents whose emotional resonance persists
through different instantiations. For responses to this, see Davies (1994) 25–6 with
bibliography.

ancient thinking is frequently directed at the consequences of music’s
effects on its listeners. Part of Plato’s response to this problem is the
subject of Anastasia-Erasmia Peponi’s essay, which traces a tension
between semantics and musical affectivity that runs through Plato’s
discussion of lyric poetry in the Republic and the Laws. Peponi argues
that Plato’s privileging of the semantic element is informed by anxiety
about the tendency of rhythmic and harmonic features of such poetry
to overwhelm the verbal. On Peponi’s reading, Plato is particularly
sensitive to lyric song’s capacity to bring about scenarios in which the
non-representational dimensions of poetic performance come to
predominate, creating atmospheres and moods that threaten the
mimetic transparency on which, on Plato’s view, the efficacy of
valid poetic communication depends. His decision, in the Laws, to
include music in the category of those representational arts that have
a prominent visual component (τέχναι εἰκασταί) can therefore be
seen as an attempt, not entirely successful in theoretical terms, to
control ‘the non-representational potential of melos’39 by subordin-
ating it to a schema in which signification and mimesis are sovereign.
Pierre Destrée tracks a very different response to the problem of
music’s autonomy in Aristotle’s thinking about ‘music for leisure’.
Rather than attempting to construct a medial hierarchy in which
music is subordinate, Aristotle argues that music can be enjoyed as an
activity pursued in leisure (πρὸς τὴν ἐν τῇ σχολῇ διαγωγήν),40 which is
an end in itself, insofar as this type of music is ‘an activity through which
our “theoretical” intelligence exercises its power’.41 Andrew Barker’s
chapter on Plutarch Quaest. conv. 7.5 focuses on similar issues in the
intersections of musical theory and performance culture in the Second
Sophistic. In discussing Plutarch’s account of the aulete’s performance,
and the responses that it occasions, Barker argues that he uses the story
to stage a dialogue between competing ways of conceptualizing musical
performance, using Platonic and Aristotelian intertexts to guide readers’
understanding of the issues at stake. This chapter exemplifies a socio-
textual approach to the representation of music that has greatly illu-
minated the interpretative conditions within which ancient listeners
operated.42 In explicating the intertextual sophistication of Plutarch’s
narrative, Barker’s reading shows that musical performances still
39
Below p. 177.
40
For discussion see Destrée (this volume) 183–4.
41 42
See p. 201. Wilson (1999) is exemplary.
12 Tom Phillips
carried a powerful ideological charge in later antiquity. The critical
terms within which such performances were conceptualized were sub-
ject to debate in ways that continued and extended the debates explored
by Peponi and Destrée.
More oblique, but no less culturally significant, relationships
between music and literature are to the fore in James Porter’s discus-
sion of sound in Hellenistic literary theory, which traces the attempts
of Cicero, Longinus, and the Hellenistic κριτικοί to theorize the
musicality of language itself, the quality of which, for the ‘euphonist’
theorists at least, was the determining criterion of literary value. On
Porter’s reading of these critics, poetically realized sound is a para-
doxical phenomenon, located in a series of concrete moments, yet
producing an overall structure that cannot be grasped in terms of
any one of these instances. As well as illustrating the vital import-
ance of a materially-focused auditory aesthetics to Hellenistic liter-
ary thought,43 Porter brings out another dimension of music’s
cultural importance by stressing ‘the common sources of literary
and rhetorical criticism in the ancient theory of music and of poetry
as music’.44
Taken as a whole, this volume emphasizes the variety of musical
effects at work in ancient poetic texts and the productive volatility of
music’s place in intellectual culture throughout antiquity. As well as
building on previous scholarship, it sketches some new directions for
research: much remains to be said about the connection between
rhythm and meaning, about the influence of music on literary criti-
cism, and the methodological problems involved in conceptualizing
the emotive effects of ancient music, to name but a few of the areas
ventured into here. Although they develop fresh insights into their
subject matter, and suggest approaches that could be developed
further, the analyses collected here do not aim at being methodo-
logically programmatic. Indeed two implications of the volume are
that ancient ‘musical texts’ and texts about music require multiple
and overlapping interpretative approaches which draw on various
intellectual disciplines, and that such approaches need to be closely
attuned to the specific texts (and contexts) that they address. Many
43
Considerations of performance are also important to Porter’s treatment of the
‘euphonist’ literary critics, whom he sees as motivated in part by the need to reanimate
performance texts that readers of later times accessed primarily through books.
44
Pp. 216.

readers will doubtless find these formulations unsurprising: stated
thus, these interpretative parameters reflect an approach to cultural
phenomena that has been prevalent across various regions of literary
and cultural criticism in recent decades, and would probably be
endorsed by most scholars as guidelines for all manner of interpret-
ative practices.
Yet this volume also suggests that ancient music poses challenges
to this model. While some texts can be productively examined with
an analysis that sees music as a handmaiden to language, there are
other moments at which musical texts require a recalibration of
established literary critical procedures. Various chapters in this vol-
ume highlight points at which ‘multiple and overlapping interpret-
ative approaches’ do not necessarily entail the development of
coherently heterogeneous readings, but serve to focus attention either
on tensions within the texts under discussion, or gaps in our means of
understanding them. Analyses such as those in the second half of the
volume demonstrate that the ancients wrestled with these tensions
and gaps just as much as modern readers, and the questions opened
up by close engagement with poetic texts suggest that more work
remains to be done in articulating modes of criticism that will do
justice to the complexity of ancient musico-poetic culture.
Partly because of the range of the material it covers, therefore, and
partly because of the varying interpretative challenges posed by its
subjects, this volume does not offer unitary, much less straightfor-
ward, answers to the questions posed at the outset. Instead, it consti-
tutes a series of interpretative gestures that both clarify and
complicate the terms on which such questions might be approached.
Readers will note that the essays employ a wide range of literary
critical, musicological, and socio-cultural modes of analyses; each
entails problems and opportunities that should stimulate further
debate. Thematic diversity is also apparent. The centrality of music
to Greek cultural life means that its influence is felt across genres,
practices, and contexts: its imagined echo in the lawcourt and
scholar’s study is no less powerful than its fully realized sound in
the theatre or symposium, and its affective range in relation to
different types of subject-matter matches the generic variety of
Greek literature itself. In tackling this variety of material and meth-
odology, the volume offers templates with which scholars across a
range of disciplines might think further about how consideration of
music could enrich their engagement with ancient culture.
Part I
Interpretation
Epicentric Tonality and the Greek

Lyric Tradition
John C. Franklin
In 1996, as a doctoral candidate at University College London, I set

out to compare Greek and Indic music terminology.1 I had learned
some basics from West’s recent Ancient Greek Music (1992), includ-
ing the early importance of the lyre and its middle string (μέση) that
served as some kind of tuning and/or tonal center. I had also encoun-
tered an old proposal by Fox-Strangways that some early Indian
ideas—including a central scale-degree madhyama and microtonal
intonations—might be cognate with those of Greece via Indo-
European inheritance.2 But I soon stumbled upon a more recent
suggestion that madhyama could rather be a relic of a Mesopotamian
tuning and/or tonal conception that featured a complete diatonic/
heptatonic tuning cycle.3 This system, reconstructed from a handful
of relevant tablets by Kilmer, Gurney, and others mainly between 1960
and 1970, went back to the Old Babylonian period (c.2000–1600) and
1
This paper derives from three talks: ‘The Epicentric Arrangement of the Archaic
Heptachord’, Ionian University, Corfu, Greece (7/2007); ‘The Middle Muse: An
Overdue Book’, Music and Text in Ancient Greece (Oxford, June 29–30, 2013); ‘East
Faces of Early Greek Music’, American Philological Association annual meeting,
MOISA panel (1/2015).
2
Fox-Strangways (1914) 142.
3
Kilmer (1971) suggested several parallels to specific Sanskrit terms which, as far
as I know, have yet to be addressed (though the possible relevance of madhyama was
noted by Widdess (1995)). The best general discussion I have found of a potential
Indo-Mesopotamian link is Picken (1975) 601ff.
18 John C. Franklin
beyond to Sumerian tradition.4 I immediately shifted my comparative
study to these texts, and eventually argued (2002b) for vestigial echoes
of the Mesopotamian system in Greek lyre practice of the Archaic
period (c.750–500), aspects of which can be deduced from later
sources.5 I also proposed an historical and cultural scenario in several
stages for this transmission and the subsequent evolution of what I saw
as a syncretism with inherited Greek musical traditions (especially
epic).6 But despite a quick contract with Oxford for a monograph
called The Middle Muse,7 further research and feedback from col-
leagues soon convinced me that my historical apparatus was inad-
equate and inaccurate at several junctures. Wishing to provide
credible context(s) that would bolster, not undermine, the technical
arguments, I began to examine various historical and cultural
‘moments’ in more detail—a process that has lasted fifteen years,
culminating in a study of divinized temple-instruments in the ancient
Near East and their imprint on Kinyras, the mythological lyre-king of
pre-Greek Cyprus.8 With this I can finally return to the technical
material with which I began—although here too significant revisions
are now necessary given recent advances in our understanding of
Greek tonal history, thanks especially to the intervening work of
Stefan Hagel.9
The present paper is a stopgap until I complete The Middle Muse,
or a summary statement should that never come to pass.10 I shall first
4
For the key texts and early reconstruction, see esp. Kilmer (1965); Gurney
(1968); Wulstan (1968); Kilmer (1971); further subsequent literature reviewed in
Kilmer (1997); to the corpus one must now add the heptatonic ‘star-text’: see
references in Franklin (2015) 41 n. 148. Sumerian background: Kilmer (1965) 261;
Shaffer (1981) 82–3; Krispijn (1990); Gurney (1994).
5
The preliminary comparison of Lasserre (1988) was marred by faulty arguments
and cursory knowledge of the literature, often relying on outdated texts and very
speculative musicological interpretations (more detailed critique in Franklin (2002b)
passim). Nevertheless Lasserre did offer some interesting suggestions, and concluded
merely with an appeal for honest reinterpretation of early Greek musical history.
6 7
Franklin (2002b). For the title, see further p. 44.
8 9
Franklin (2015). Hagel (2009).
10
In the latter bittersweet event, Chapters 1, 6, and 8–10 of Franklin (2002b)
contain technical interpretative arguments not developed elsewhere, and would still
be worth consulting (despite many adjustments I would now make). The remaining
chapters are now largely obsolete; but since they are unfortunately published online
(by UCL), I would redirect potential readers to later papers that elaborated, refined,
and corrected ideas worth keeping. For Chapter 2, see Franklin (2002c); Franklin
(2008); Franklin (2012) with historical modifications sketched in the present paper;
Chapter 3, Franklin (2004) and Franklin (2011); Chapter 4, Franklin (2006) and
Epicentric Tonality and the Greek Lyric Tradition 19

present the key primary texts and arguments that I believe make some
connection between the early Greek lyre tradition and the Mesopo-
tamian tonal/tuning system unavoidable.11 I shall then briefly sketch
the evolution of my approach to the historical situation, surveying the
contextual studies alluded to above and offering new suggestions.
A closing discussion will relate the ‘Middle Muse’ to this volume’s
central topic—music and text.
EPICENTRIC TONALITY BETWEEN GREECE

AND MESOPOTAMIA
The discovery of a very early diatonic tuning/tonal system in Mesopo-

tamia fatally undermined previous a priori evolutionary assumptions
about Greek tonal history. Winnington-Ingram’s Mode in Ancient
Greek Music (1936) had established the belief—still accepted by West,
Barker, and others when I began—that the earliest tunings of which we
hear were ‘defective’, the Greeks not ‘achieving’ until the late fifth or
fourth century the complete diatonic conception that underlies the
Perfect System and the Aristoxenian pitch keys (τόνοι).12 But the
Mesopotamian texts now showed that a complete diatonic cycle had
been elaborated 1500 years and more earlier. This comprised seven
diatonic/heptatonic tunings that could be generated by tuning string-
pairs to resonant/consonant13 ‘fifths’ (3:2) and ‘fourths’ (4:3), and
knowledge of how to convert one tuning to the next by ‘clearing’ the
unique tritone in each (that is, adjusting one of its strings to another fifth
or fourth, thus ‘displacing’ the tritone to a new location). It is essential to
realize that, while the tablets may give the impression of systematic
Franklin (2015); Chapter 5, Franklin (2003); Chapter 7, Franklin (2002d); Franklin

(2002c) 446–9; Franklin (2003) 303–6; Franklin (2005) 13–22.
11
It is not possible here to discuss fully all relevant textual and interpretative issues
involving these tablets. For my own assumptions, if not otherwise clarified here, please
see Franklin (2002b) especially Chapter 6.
12
Winnington-Ingram (1936); cf. Franklin (2002d) 669–72.
13
‘Consonant’ is derived from Latin consonans, a calque for Greek σύμφωνος. By a
convention going back to Aristoxenus, the word was applied only to the 3:2 ‘fifth’ and
4:3 ‘fourth’. But these stand at the head of a continuum of such relationships (5:4, 6:5,
7:6, and so on), for which the more general ‘resonant’ is emerging as a scholarly term:
see Crocker (1997); Franklin (2005) 12–13; Hagel (2009) passim.
20 John C. Franklin
theoretical elaboration, the whole cycle is derived quite naturally from a
few definite physical/auditory phenomena, each with its own unique
sonic identity (Figure 1.1).14 The experience of sound was fundamental,
as we are reminded by the Greek terms for consonance, σύμφωνος
(‘speaking together’) and the early alternative συνᾷδον (‘singing
together’, Heraclit. 10 D-K, etc.).15 In Greece too, I argued, diatonic
lyre tunings, again created through constructing consonant intervals—
what Aristoxenus calls ‘taking through consonance’ (ἡ λῆψις διὰ
συμφωνίας),16 though the aural experience was expressed from earliest
times by a metaphor of ‘joinery’ (ἁρμονία and relations)17—must have
been much older than generally believed.18 First, Aristoxenus, drawing
on a musicians’ tradition, explicitly states that the diatonic was the
oldest and most natural of the γένη.19 Second, his ‘first principle’
14
For Greek perceptions of the diatonic as the ‘most natural’ tuning, see Franklin
(2002d) 672–3.
15
For stimulating observations on the primacy of sound, see Crocker (1997)—in
my view a seminal methodological essay despite the scathing response of Gurney/
West (1998). Note that in both Greek and English the terms ‘fourth’ (ἡ [sc. συμφωνία]
διὰ τεττάρων) and ‘fifth’ (ἡ διὰ πέντε) involve a revealing hysteron proteron, referring
as they do to the number of scale degrees comprised by each when the process of
tuning by ‘fifths’ and ‘fourths’ is taken to its ultimate, diatonic/heptatonic conclusion.
But these intervals have an aural identity anterior to and independent of that process
(Figure 1.1). There is no Akkadian equivalent for ‘fifth’ and ‘fourth’ (each unique pair
of strings standing three, four, five, or six strings apart, counting inclusively, has its
own name in CBS 10996 and the Retuning Text UET VII 74: Crocker (1978)).
16
For ἡ λῆψις διὰ συμφωνίας, see Aristox. Harm. 55–6; cf. Eucl. Sect. Can. 17
(162.1ff.); Ps.-Plut. De mus. 38.1145b–c; Ptol. Harm. 1.16 (39.14ff.); 2.1 (44.1ff.); 2.9
(62.1ff.); Aristid. Quint. 2.14 (80.2–3).
17
For ἁρμονία etc., see further pp. 31–2. Ζεύγνυμι and its relations are also relevant,
e.g. διάξευξις (‘disjunction’ of tetrachords) and even the lyre’s ζυγόν, since the resonant/
consonant tuning of strings, arrayed along the yoke, constituted a further aspect of the
instrument’s physical ‘harmony’; cf. the Homeric Hymn to Hermes 50–1 (καὶ πήχεις
ἐνέθηκ’, ἐπὶ δὲ ζυγὸν ἤραρεν ἀμφοῖν, / ἑπτὰ δὲ συμφωνοὺς ὀίων ἐτανύσσατο χορδάς);
Aristophanes’ description of the old, drunken Cratinus as a worn-out, slack-stringed lyre
with gaping joints/‘harmonies’ (ἐκπιπτουσῶν τῶν ἠλέκτρων καὶ τοῦ τόνου οὐκέτ’ ἐνόντος
/ τῶν θ’ ἁρμονιῶν διαχασκουσῶν, Eq. 532–3); and probably Timoth. Pers. fr. 15.224–225
(PMG 791), of Terpander. Διάτονος itself may accord with the construction metaphor, as
the word, along with διατόν(α)ιον, was also used of ‘bonding courses in a wall’, joists, and
the like; cf. LSJ s.vv. with references, adding Inscriptions de Délos I/3 290 and III/5 1417.
Carpentry also seems to underlie Akk. pitnu, for which see Kilmer (1965) 262–5; Kilmer
(1971) 132; Franklin (2002d) 677.
18
Further technical argument: Franklin (2002c); Franklin (2002d); Franklin
(2003) 303–6; Franklin (2005) 13–22.
19
Aristox. Harm. 19: πρῶτον μὲν οὖν καὶ πρεσβύτατον αὐτῶν θετέον τὸ
διάτονον, πρῶτον γὰρ αὐτοῦ ἡ τοῦ ἀνθρώπου φύσις προστυγχάνει, δεύτερον δὲ τὸ
χρωματικόν, τρίτον δὲ καὶ ἀνώτατον τὸ ἐναρμόνιον, τελευταίῳ γὰρ αὐτῷ καὶ μόλις
1:1 6:5
3:2 7:6
4:3 8:7
5:4 9:8
Figure 1.1. Oscilloscope readings of two tones at unison, the 3:2 ‘fifth’, the
4:3 ‘fourth’, and so on through the 9:8 wholetone (Gk τόνος).
(ἀρχή) of ‘continuity’ or ‘cohesion’ (συνέχεια) required that a properly

formulated heptatonic scale (μέλος ἡρμοσμένον) conform to minimum
diatonic conditions by having every note respond by a consonant fourth
and/or fifth to a note four or five degrees away; since this principle is
most consistently valid with diatonic tuning, the rule must be derived
from diatony, which therefore stands in relation to the chromatic and
enharmonic as genus to species (even if Aristoxenian theory itself used
μετὰ πολλοῦ πόνου συνεθίζεται ἡ αἴσθησις (‘Now, the diatonic must be put down as the
first and oldest of them [sc. the genera], for the nature of man comes across it first, and
afterwards the chromatic, and third and finally the enharmonic, for it is the last to
which the perception grows accustomed—and with difficulty at that, after much
labor’). That this idea is not merely a priori speculation (Nagy 1990: 100–1) is
shown by Aristoxenos’ appeal to musical tradition (ὑπολαμβάνεται ὑπὸ τῶν μουσικῶν)
in deriving the enharmonic of Olympos from the diatonic (fr. 83 = Ps.-Plut. De mus.
1134f–1135b). See further Franklin (2002d) 672–3 (with other sources for the diatonic
as ‘most natural’), 691 (Olympos); Franklin (2002c) 447–9.
22 John C. Franklin
the term γένος for all three).20 To put it another way, the ‘family’ of
which the three γένη are parallel offspring is heptatony; but diatony is
the heptatony par excellence and imposes its ‘continuous/cohesive’
structure upon the other two γένη. Συνέχεια must be understood against
the unambiguous literary and iconographical evidence that lyres were
normally equipped with seven strings in the Archaic period, with
Mycenaean and even Minoan antecedents.21
While diatonic tuning may seem natural enough to have arisen
independently in Greece and Mesopotamia, further parallels are, in
my view, too specific to be coincidence. These are fourfold, and reveal
a consistent theoretical and practical perspective that I would call
epicentric tonality. On the Mesopotamian side (Figure 1.2):
1. The strings are named and numbered in relation to a central
pitch, that is 1–2–3–4–5–4b–3b–2b–1b (where ‘b’ reflects the
qualification ‘of the back strings’).22
20
Aristox. Harm. 29: ὑποκείσθω δὲ καὶ τῶν ἑξῆς κειμένων φθόγγων κατὰ μέλος ἐν
ἑκάστῳ γένει ἤτοι τοὺς τετάρτους [τοῖς τετράσι] διὰ τεττάρων συμφωνεῖν ἢ τοὺς
πέμπτους [τοῖς πέντε] διὰ πέντε ἢ ἀμφοτέρως (‘And let it also be laid down that, for
notes which are “continuous” along a scale [μέλος, sc. ἡρμοσμένον]—in each genus—
either every fourth note is consonant at a fourth, or every fifth note is consonant at a
fifth, or both’). Harm. 54: οὐ δεῖ δ’ ἀγνοεῖν, ὅτι οὐκ ἔστιν αὔταρκες τὸ εἰρημένον πρὸς
τὸ ἐμμελῶς συγκεῖσθαι τὰ συστήματα ἐκ τῶν διαστημάτων· οὐδὲν γὰρ κωλύει
συμφωνούντων τῶν φθόγγων κατὰ τοὺς εἰρημένους ἀριθμοὺς ἐκμελῶς τὰ συστήματα
συνιστάναι, ἀλλὰ τούτου μὴ ὑπαρχόντος οὐδὲν ἔτι γίγνεται τῶν λοιπῶν ὄφελος. θετέον
οὖν τοῦτο πρῶτον εἰς ἀρχῆς τάξιν οὗ μὴ ὑπαρχόντος ἀναιρεῖται τὸ ἡρμοσμένον (‘It is
essential to realize that the aforementioned [sc. principle] does not guarantee that
systems will be properly assembled from intervals. For nothing stops a tuning from
being put together improperly even when the notes are consonant according to the
aforementioned numbers [i.e. every note being consonant by a fourth or fifth or
both with every fourth or fifth note from itself]; but if this condition is not fulfilled,
there is no use bothering about the rest: and so this must be made the first principle,
without the fulfilment of which the attuned scale [τὸ ἡρμοσμένον, sc. μέλος] is
destroyed’). See further Franklin (2002c), esp. 446–7; cf. Franklin (2002d) 670,
673; Franklin (2005) 19; as prefigured in the Homeric Hymn to Hermes 51–2, see
Franklin (2003) 303–6.
21
For literary sources, see Franklin (2002d) 686 and notes 44–5, 47; Franklin
(2005). For the issue of heptatonic continuity in the Aegean from the Late Bronze Age,
see further pp. 27–35.
22
This is most clearly seen in the text U.3011 which, though a Neo-Babylonian
tablet from Ur, was shown to reflect much older tradition by the duplicate fragment
N.4782 (Nippur, c.1800–1500): Kilmer (1960); Kilmer (1965); official publication as
UET VII 126 (Gurney (1974); Finkel/Civil (1982); cf. Kilmer (1997).

TUNING: qablitu 5–2 nis̆ gabarî 1–5 nid qabli 4–1 pitu 3b–4 embubu 3 –3b kitmu 4b–3 is̆artu 2–4b
UNCLEAR: nis̆ gabarî 1(2b)–5 nid qabli 4–1(2b) pitu 3b–4 embubu 3–3b kitmu 4b –3 is̆artu 2(1b)–4b
TIGHTEN: 1 and 2b 4 3b 3 4b 2 and 1b
higher
1 1 1 1 1 1
1
TIGHTENING
2
2 2 2 2 2 2 2
3 3 3
3 3 3 3
pitch in semi-tones
4 4 4 4 4
4 4
5 5 5 5 5 5 5 5
4b 4b
4b 4b 4b 4b 4b
3b 3b 3b 3b
3b 3b 3b
2b 2b 2b 2b 2b 2b
2b 1b
lower
1b 1b 1b 1b 1b 1b
TUNING: is̆artu 2–4b kitmu 4b–3 embubu 3–3b pitu 3b–4 nid qabli 4–1 nis̆ gabarî 1–5 qablitu
UNCLEAR: qablitu 5–2(1b) is̆artu 2(1b)–4b kitmu 4b–3 embubu 3–3b pitu 3b–4 nid qabli 4 –1(2b)
LOOSEN: 2 and 1b 4b 3 3b 4 1 and 2b
higher
1 1 1 1 1 1
2 1
LOOSENING
2 2 2 2 2 2
3 3 3
3 3 3 3
pitch in semi-tones
4 4 4 4 4
4 4
5 5 5 5 5 5 5
4b 4b
4b 4b 4b 4b 4b
3b 3b 3b 3b
3b 3b 3b
2b 2b 2b 2b 2b 2b
1b 2b
lower
1b 1b 1b 1b 1b 1b
Figure 1.2. Graphic representation of the Mesopotamian tuning cycle UET

VII 74.
2. A sequence of seven diatonic tunings, attested by the ‘Retuning

Text’ (UET VII 74, Ur, c.2000–1600) and several others,23
begins from one called ‘middle’, which is itself tuned beginning
from the middle string.24
3. This middle string is the only one which does not change in
pitch throughout the complete heptatonic/diatonic cycle.25
23
The seven tunings’ order is also found in a Middle Assyrian (c.1100) song-
catalogue from Assur (VAT 10101 col. viii.45–52), which matched the surviving
sequence in the Retuning Text and permitted the latter’s restoration. For this tablet,
see Ebeling (1919) no. 158; Ebeling (1922); Kilmer (1965) 267; Kilmer (1971) 138;
Bayer (2014) 24, 32–5.
24
Kümmel (1970) perceived the crucial relationship between the tuning names and
the named string/interval from which each tuning is generated via ‘taking through
consonance’ (at least theoretically, that is, within the system’s own nomenclature).
25
This assertion remains valid even if the Retuning Text is to be restored beyond the
seven distinct tunings, which would then require the central string to be raised or
lowered (for this issue, see Gurney (1968) 232–3; Wulstan (1968) 221; Duchesne-
Guillemin (1969) 12; Gurney (1994), 102ff.). For it would remain the case that the
central string does not move across the seven standard tunings, which inevitably
constitute a self-contained system, regardless of any further uses to which they were put.
24 John C. Franklin
4. In the one largely intact specimen of the Hurrian hymns from
thirteenth-century Ugarit, which use a local form of the Meso-
potamian (Akkadian) interval names somehow to map out
harmonic progressions, the middle string (5) and two others
(2, 4b) connected to it by a direct chain of consonance featured
rather more prominently than the others.26 This observation is
crucial for establishing that the epicentric structure/nomencla-
ture of strings, intervals, and tunings was no mere artifact of
theory and/or scribal tradition, but had some bearing on per-
formance.27 This deduction is corroborated by the Hurrianized
form of the Akkadian terms, which indicates a considerable
period of oral transmission;28 and the fact that this collection
26
On the ‘downwards’ reading that is now generally accepted, i.e. with string one
as highest in pitch, the relevant pitches of the nid qabli tuning may be represented as
3b(F)-4b(G)-5(A)-4(B)-3(c)-2(d)-1(e), with lowercase letters representing a higher
octave (it is worth pointing out, however, that on the old ‘upwards’ reading, strings
1-2-3-4-5-4b-3b in qablitu tuning would yield the conjunct heptachord of Greek
theory). For a thorough and subtle statistical analysis of interval frequency in the
Hurrian hymns and the modal/tonal implications of interval sequences, see Hagel
2005b (with references in 297 n. 27 for the ‘downward’ interpretation). My own more
Cromagnon count of interval and string frequency, prepared for a lecture in February
2003 (at CAARI), yielded the following figures (those marked with * leave out 10
ishartu as a possible scribal error [the figure may seem anomalously high]; strings 2b
and 1b are not separately counted, as being octave-repetitions of 1 and 2 [as seen in
the Retuning Text and CBS 10996]). Interval Frequency with percentage of attested
total: qablitu (2-5): 3+3+3+4+2 = 15 = 20.82% shalshatum (1-6): 2+2+4+4+2+2=16
= 22.21% ishartu (2-6): 10 = 13.88% seru (7-5): 1+2+2+1+1+1=8 = 11.10%
rebutu (2-7): 1+2+1+1+2 = 7 = 9.72% titur ishartum (3-5): 2+4 = 6 = 8.33% kitme
(6-3): 2+1+1 = 4 = 5.55%. sirdu (4-6): 1+1 = 2 = 2.78% nid qabli (4-1): 1+1 = 2 =
2.78% embubu (3-7): 1 = 1.39% titur qabli (2-4): 1 = 1.39%. String Frequency with
percentage of attested total (derived from Interval Frequency: each interval consists of
two strings, but the same string can appear in more than one interval): 2: 15+7+[10]
+1 = 33 (*23) = 22.92% (*18.55%) 5: 15+8+6 = 29 = 20.14% (*23.70%) 4b [10]+2
+16+4 = 32 (*22) = 22.22% (*17.74%) 3b: 7+8+1 = 16 = 11.11% (*12.90%) 1: 16+2
= 18 = 12.5% (*14.51%) 3: 6+1+4 = 11 = 7.64% (*8.87%) 4: 2+2+1 = 5 = 3.47%
(*4.03%). Note the infrequency of nid qabli, though that is the interval from which
one would (theoretically) begin to achieve this tuning. Instead, the most prominent
(‘repeated’) strings are 5, 2, and 4b—a set that may be compared with the common
disjunctive tunings of Greek tradition, where the central string (μέση) is separated by
a whole-tone from its neighbour παραμέση, as are strings 5 and 4b in nid qabli tuning;
similarly 2 and 5 are separated by a fourth, as are μέση and ὑπάτη, the lower boundary
of early Greek lyre-tuning. In other words, the tones of the consonant framework are
emphasized, as is not surprising.
27
For the Ugaritian hymns, see esp. Hagel (2005b).
28
Hagel (2005b) 293 n. 22.

of hymns was archived according to the practical criterion of
tuning, and therefore saw active liturgical use at Ugarit.29
These features find the following counterparts in the early Greek
lyre tradition:
1. an epicentric arrangement inherent in the string names them-
selves, most explicitly acknowledged by Aristotle (Figure 1.3),30
but equally implicit in the use of ‘highest, middle, and last’, by
Plato and others, as a shorthand for the complete stringing of
the lyre;31
2. statements in the Aristotelian Problems and later Dio Chrysostom
that the middle string (μέση) was the first to be tuned and the
others were tuned to it (for the Koine Hormasia, see p. 41);32
3. that μέση was the only string that, if its pitch were changed after
the instrument was tuned, would spoil the overall attunement
(τὸ ἡρμόσθαι);33
29
The recovered hymns, so far as we know, were all in the nid qabli tuning. The
same organizing principle is seen in the song catalogue VAT 10101 mentioned above.
Cf. Franklin (2015) 97 (where qablītu in n. 47 is a lapse).
30
Arist. Metaph. 4.1018b26–29: τὰ δὲ κατὰ τάξιν (ταῦτα δ’ ἐστὶν ὅσα πρός τι ἓν
ὡρισμένον διέστηκε κατά τινα λόγον, οἷον παραστάτης τριτοστάτου πρότερον καὶ
παρανήτη νήτης· ἔνθα μὲν γὰρ ὁ κορυφαῖος ἔνθα δὲ ἡ μέση ἀρχή (‘Other things
[sc. are called prior or posterior] with respect to arrangement. These are whatever
things stand at intervals according to some numbering with reference to some defined
point. For instance, the second-man-in-line is before the third-man-in-line, and
παρανήτη is before νήτη: in the one case the chorus-leader is the starting point
(ἀρχή), in the other the middle-string (μέση)’. My diagram favours παραμέση over
τρίτη: for this issue, see n. 73.
31
Pl. Rep. 4.443d: ὥσπερ ὅρους τρεῖς ἁρμονίας ἀτεχνῶς, νεάτης τε καὶ ὑπάτης καὶ
μέσης, καὶ εἰ ἄλλα ἄττα μεταξὺ τυγχάνει ὄντα (‘Just as three boundaries of tuning
(ἁρμονία)—νεάτη, ὑπάτη, and μέση—and whatever others happen to be between
them’). For this trope, see further Zaminer (1984).
32
Dio Chrys. 68.7: ὥσπερ ἐν λύρᾳ τὸν μέσον φθόγγον καταστήσαντες ἔπειτα πρὸς
τοῦτον ἁρμόζονται τοὺς ἄλλους· εἰ δὲ μή, οὐδεμίαν οὐδέποτε ἁρμονίαν ἀποδείξουσιν κτλ.
(‘And as in the lyre, one must establish the middle tone, and then tune the others to it;
otherwise, they will never display any ἁρμονία’). For the relevant Aristotelian Prob-
lems, see nn. 33 and 34.
33
Ps.-Arist. Prob. 19.36: Διὰ τί, ἐὰν μὲν ἡ μέση κινηθῇ, καὶ αἱ ἄλλαι χορδαὶ ἠχοῦσι
φθεγγόμεναι, ἐὰν δὲ αὖ ἡ μὲν μένῃ τῶν δ’ ἄλλων τις κινηθῇ, κινηθεῖσα μόνη φθέγγεται; ἢ
ὅτι τὸ ἡρμόσθαι ἐστὶν ἁπάσαις, τὸ δὲ ἔχειν πως πρὸς τὴν μέσην ἁπάσαις, καὶ ἡ τάξις ἡ
ἑκάστης ἤδη δι’ ἐκείνην; ἀρθέντος οὖν τοῦ αἰτίου τοῦ ἡρμόσθαι καὶ τοῦ συνέχοντος
οὐκέτι ὁμοίως φαίνεται ὑπάρχειν. μιᾶς δὲ ἀναρμόστου οὔσης, τῆς δὲ μέσης μενούσης,
εὐλόγως τὸ κατ’ αὐτὴν ἐκλειπόμενον· ταῖς γὰρ ἄλλαις ὑπάρχει τὸ ἡρμόσθαι (‘Why is it
that, if μέση is changed, the other strings also sound spoiled, whereas if μέση remains
while one of the other strings is changed, only the changed string is spoiled? Is it
26 John C. Franklin
μέση
middle-finger string
λιχανός παραμέση
licking-finger string next-to-middle string
παρυπάτη παρανήτη
next-to-top string next-to-last string
ὑπάτη νήτη
top string last string
Figure 1.3. Epicentric arrangement of traditional Greek heptachord, according

to Aristotle.
Drawn by Bo Lawergren and originally published in B. Lawergren (1998) ‘Distinctions among
Canaanite, Philistine, and Israelite Lyres, and their Global Lyrical Contexts’, Bulletin of the
American Schools of Oriental Research 309, 41–68.
4. that μέση was often repeated in ‘worthy melodies’ (τὰ χρηστὰ

μέλη)34—a loaded expression invoking traditional practice
anterior to the fifth-century New Music.35
because for all the strings being in tune consists of having some relation towards
μέση—and the pitch of each is already [sc. established] through that string. Thus,
when you take away the cause of their being-in-tune and that which holds them
together, it no longer appears to be the same. But if one of the strings is out of tune
while μέση maintains its pitch, it makes sense for that string alone to be left out of the
tuning, since the being-in-tune persists for the others’).
34
Ps.-Arist. Prob. 19.20: Διὰ τί, ἐὰν μέν τις τὴν μέσην κινήσῃ ἡμῶν, ἁρμόσας τὰς
ἄλλας χορδάς, καὶ χρῆται τῷ ὀργάνῳ, οὐ μόνον ὅταν κατὰ τὸν τῆς μέσης γένηται
φθόγγον, λυπεῖ καὶ φαίνεται ἀνάρμοστον, ἀλλὰ καὶ κατὰ τὴν ἄλλην μελῳδίαν· ἐὰν δὲ
τὴν λιχανὸν ἤ τινα ἄλλον φθόγγον, τότε φαίνεται διαφέ ρειν μόνον, ὅταν κἀκείνῃ τις
χρῆται; ἢ εὐλόγως τοῦτο συμ βαίνει; πάντα γὰρ τὰ χρηστὰ μέλη πολλάκις τῇ μέσῃ
χρῆται, καὶ πάντες οἱ ἀγαθοὶ ποιηταὶ πυκνὰ πρὸς τὴν μέσην ἀπαντῶσι, κἂν ἀπέλθωσι,
ταχὺ ἐπανέρχονται, πρὸς δὲ ἄλλην οὕτως οὐδεμίαν (‘Why is it that, if someone moves
μέση, after tuning the other strings, and uses the instrument, it grates and sounds out
of tune, not only when it comes to μέση, but also during the rest of the melody; yet if
someone changes λιχανός or some other note, then the instrument appears to be out of
tune only when someone uses that string? Is this only to be expected? For all good
melodies make frequent use of μέση, and all the good composers . . . if they depart
from μέση, quickly return to it, as they do to no other string’).
35
For this part of the historical framework, see Franklin (2013) with further
references. The surviving Greek technical tradition relates mainly to developments
originating with the New Music: Hagel (2009) as a whole. For the movement more

While these parallel features are certainly not identical, I believe
they are sufficiently close, and too particular, to be dismissed as
coincidence. Indeed the hypothesis is made more believable precisely
by several major differences—the lack of any counterpart to the
Akkadian interval names, the systematic dichordal analysis to which
they were attached, and the nine-stringed Sumerian expression of the
heptatonic tonal phenomena versus the early Greek insistence on
seven, that magical minimum36—since one should hardly expect
complete correspondence between two musical cultures separated
by so many miles and centuries. We must allow for considerable
vagaries of transmission and tradition in several cultural/historical
stages, within what must have been largely oral/aural environments.37
HISTORICAL SKETCH
Anyone who accepts some connection between the Mesopotamian

tuning/tonal system and the early Greek lyric tradition is left pon-
dering the historical and cultural circumstances that might account
for this startling phenomenon—not to mention its implications for
literary and mythological parallels between the two cultural spheres,
especially those relating to or deriving from cult traditions.
generally, including socio-political and literary dimensions, Richter (1968); Wilson

(1999); Wilson (2003); Csapo (2004); Power (2010) 500–54; D’Angour (2011) 202–6;
LeVen (2014).
36
The Retuning Text (in which strings 1 and 2 are doubled at the octave by 2b and
1b) and CBS 10996 (which features only strings 1–2–3–4–5–4b–3b and assigns them
numbers between 1–7) show that the heptatonic nature of the tonal phenomena was
clearly apprehended. Moreover the Akkadian name of string 4, ‘Ea-Creator’, being the
only one distinguished by a divine epithet, seems to emphasize its position within a
stretch of seven strings, not nine (Franklin 2006: 63). Carefully rendered seven-
stringed harps in terracotta plaques of Old Babylonian date (Rashid 1984: 80–8 and
figs. 62–4, 67, 72–5) indicate that, around when the Retuning Text was composed and/
or copied, musicians were well aware that seven strings were sufficient for effecting all
tuning phenomena of the diatonic cycle. Seven is indeed the minimum number for
doing so. Considerations of ‘economy’ might thus account in part for the seven-
stringed lyres of the Mycenaean and Archaic periods. But we must remember that the
number seven enjoyed a kind of prestige in the Aegean lyric tradition, with clear Near
Eastern parallels: Franklin (2006) 54–63; Franklin (2015) 40–1 and passim.
37
For the oral/aural dimension of cult lyric at Ugarit, in which the Hurrian hymns
with notation must have seen active liturgical use, see Franklin (2015) 97, 119, 448–9.
28 John C. Franklin
One must appreciate at the outset that much of the key Greek
evidence either pre-dates Aristoxenus (our earliest substantially
extant technical authority) or comes from sources like Plato, Aris-
totle, the Aristotelian Problems, and later Pythagoreanizing writers
that perpetuate older traditional conventions against the more pro-
fessional elaborations that began with the New Music in the later fifth
century. In other words, we are dealing with an early stratum of Greek
lyric practice that provided the foundation for, and persisted on the
margins of, the complex apparatus of keys and notation with which
the extant theorists are mainly concerned, and which was largely
stimulated by advances in aulos music and design.38
Under the spell of Burkert,39 I originally looked to the Orientalizing
period (c.750–650) as a likely time for Greek exposure to the Mesopota-
mian diatonic/heptatonic system, and focused on a well-known fragment
attributed to Terpander, the historical lyre-singer of seventh-century
Lesbos who—as often with Archaic poets—became a magnet for various
legends and traditional verse.40 The couplet is evidently a professional
transitional formula between two types of singing, addressed to an
unnamed god—Apollo is an attractive candidate, but the verses may be
usefully non-committal—for generic use in agonistic contexts:
σοὶ δ’ ἡμεῖς τετράγηρυν ἀποστέρξαντες (v.l. ἀποστρέψαντες) ἀοιδὴν /
ἑπτατόνῳ φόρμιγγι νέους κελαδήσομεν ὕμνους (fr. 4 Gostoli).
Putting aside four-voiced song, we shall sing / for you new hymns on
the heptatonic phorminx.
West, building on L. Deubner, had used this fragment to support his
theory of a four-stringed phorminx as the normal instrument of epic
singers.41 Accepting West’s basic premise that τετράγηρυς ἀοιδή
referred to ‘epic melodization’,42 and the ancient interpretation43
that ἑπτάτονος φόρμιγξ marked an innovation by ‘Terpander’ (or his
age), I argued that the couplet reflected from and epitomized a real
historical development—Greek exposure to a tradition of heptatonic/
38
See now especially Hagel (2009); cf. Franklin (2002d) 673–93; Franklin (2013).
39
Burkert (1992).
40
For Terpander generally, see Gostoli (1990) (Introduction); Power (2010) (Part
III). Traditional attributions: Beecroft (2008).
41
Deubner (1929); Deubner (1930); West (1981).
42
For this generic/terminological paradox, see Franklin (2004) 244–5; Franklin
(2011) 532.
43
Str. 13.2.4; Cleonid. 12 (MSG 202.8ff.)

diatonic tonality deriving from Mesopotamia. But there was an imme-
diate complication, since seven-stringed lyres were also regularly found
in Minoan and Mycenaean iconography. Attempting to harmonize all
this evidence, I suggested that diatonic/heptatonic music, presumably
on a Mesopotamian model, had first arrived to the Aegean in the Late
Bronze Age, with its far-flung palace networks and exchanges of cultic
personnel (since the Minoan and Mycenaean centres did form the
western reach of this international economy).44 The foreign artform
disappeared, I proposed—like literacy—with the collapse of Myce-
naean palace culture, leaving the field to a parallel, inherited tradition
of epic lyre-singing (and non-heptatonic instruments) like that envi-
sioned by West. The knowledge of diatony/heptatony was then
re-imported (on the hypothesis) in the Orientalizing period, and the
development was later assigned by popular memory to the famous
Terpander as πρῶτος εὑρετής.
But I soon became dissatisfied with this double-importation
scheme. Following S. Morris and others who, at that time, privileged
Levantine avenues of influence over the so-called via Anatolica,45
I had somewhat perversely downplayed Terpander’s attested connec-
tions with Lydia, looking further east to the Neo-Assyrian period, and
with some vague idea that Cyprus might somehow be important. In
my viva, however, Burkert emphasized Lydia’s position at the ter-
minus of the Persian royal road (itself going back to Neo-Assyrian
and ultimately Hittite times). So I began investigating what I came to
see as a well-defined Greco-Lydian musical movement, with Lesbos
playing a pivotal role.46 Key features, I argued, were Lydia’s known
status as an Assyrian client state in the seventh century and its
emulation of/response to elite Assyrian culture, a phenomenon other-
wise well attested on the Empire’s periphery. The most conspicuous
sign of Lydia’s participation in what Sargon called ‘A Feast of
Music’47—an active blending of regional music traditions as part of
a more complex cultural program under the pax Assyriaca—is its use
of harps in military ensembles,48 the only contemporary parallel for
which is Assyrian practice. These data are to be coordinated with
44
Liverani (1990) remains a good introduction.
45
Morris (1992). For the revival of interest in the ‘Anatolian route’, see e.g. papers
in Collins et al. (2008), and now the grand synthesis of Bachvarova (2016).
46
Franklin (2008).
47
For the expression, see references in Franklin (2008) 201 n. 3.
48
Hdt. 1.17.
30 John C. Franklin
Archaic Greek emulation of the Lydian high life (ἁβροσύνη is a key
word here) including a sudden vogue for harps, often mentioned by
the Archaic poets; note especially the tradition that Terpander mod-
eled his βάρβιτος (the tenor lyre often used in the symposium) on the
Lydian harp, called πακτίς by some Greeks.49 Also deriving largely
from this movement, I proposed, is the custom of reclining that the
Greek symposium adopted in the Archaic period.50
Despite pinpointing a specific moment of Greco-Mesopotamian
musical contact, well defined geographically and chronologically—
one in which, moreover, Terpander himself was implicated—I had
begun to doubt that heptatonic instruments really died out at the
end of the Bronze Age. First, a well-known shard from Chania
(Crete) came from a post-palatial level, indicating that the hepta-
tonic art could survive independently of the palaces.51 One obvious
conduit here is religious lyric, given the known persistence of
various cults from the Mycenaean period (I would now see the
famous Lesbian singers like Terpander, for example, as one eventual
outcome of the process: see p. 34). Second, the earliest Archaic
examples of seven-stringed lyres were found in areas of Bronze
Age continuity and diaspora, for instance Athens and Smyrna.
There were, besides, rich mythological traditions surrounding the
seven-stringed lyre and its divine and semi-divine practitioners (Her-
mes, Apollo, Amphion, Cadmus, Orpheus, Linus). These too, like the
Greek myth-cycles overall,52 exhibit regional patterns consistent with
Bronze Age roots and Dark Age continuity. In a preliminary collection
I first encountered the rich—and very early (third–second millen-
nium)—Near Eastern evidence for the divinization of chordophones
and other musical instruments, including the kinnāru of Syro-Levantine
tradition (whence Kinyras of Cyprus and, indirectly and later, King
David).53 These were often associated with royal cult and the symbolic
development of music in royal ideology and ritual poetics.54 These
49
Pind. fr. 125.
50
This idea has found approval in two recent studies of the Archaic symposium:
Baughan (2013); Węcowski (2014).
51
Chania XM 2308, Late Minoan III: Maas/Snyder 1989, 2, 16, and fig. 2b. I owe
this observation to Stefan Hagel (Summer 2002).
52 53
Nilsson (1932). Franklin (2006).
54
Franklin (2015) (for comments about the [probably vital] relationship between
divinized chordophones and the mobility of the Mesopotamian tuning system, see
40–1, 59, and 171), with Heimpel (2015).

points, I believe, cement the Late Bronze Age as the most viable single
period for the importation of the Mesopotamian diatonic/heptatonic
cycle in a form recognizably akin to that found in the tablets. The
hypothesis can be further supported, in a general way, by abundant
evidence for a musical aspect to the artisan mobility that is otherwise
well documented for the Bronze Age palaces.55 More specific support
comes from dialectal variants of the string-names, implied or attested
for both Thebes56 and Argos57; these, when combined with the dom-
inant Attic-Ionic forms of the musiocographers, should suffice to estab-
lish some form of second-millennium inheritance. It is essential,
I believe, that the word ἁρμονία derives from—indeed preserves—a
Mycenaean form, as shown by the sonant nasal’s realization as -on-,
not -an-.58 It may even be that ἁρμονία, in the broad sense of ‘musical
tuning’ (versus a specific ἁρμονία that results), once implied a cyclical
conception of diatony/heptatony, given that Mycenaean a-mo (eventu-
ally /harmon/), ‘a thing joined’—a word that is structurally anterior to
ἁρμονία—had in that period developed the specialized and prestigious
sense ‘wheel’. Ἁρμονία might thus have been the ‘musical wheel’, each
‘turning’ of which (τρόπος?) generated a specific ἁρμονία.
My new conviction that the Greek diatonic/heptatonic tradition
reached back to the Bronze Age was strengthened by two studies of a
55
Franklin (2007); Franklin (2015), especially Part One.
56
Aeolic variants are implied by two of the legendary gate-names of seven-gated
Thebes, rendered by various non-Theban authors (going back to Aeschylus and
Euripides; Berman: 2007) as νήϊται or νήϊσται and ὕψισται (canonical νήτη and
ὑπάτη). Cf. Arcadian νήατος, LSJ s.v. νέατος; Hesych. s.v. νήϊστα· ἔσχατα, κατώτατα
(which makes better sense of string position than gates in a circle).
57
A third-century BC dedication from Argos (SEG 30.382; Kritzas: 1980) mysteri-
ously but definitely (pace Mojsik 2011: 68 n. 2) lists the three boundary strings that
epitomize the epicentric nomenclature, along with another called ‘first’ (A: Νήτας /
Μέσσας B: Ὑπάτας / Πράτας. For Πράτα, see further p. 40 n. 98). There is no reason to
assume that these names were reverse-engineered from the Attic-Ionic forms; rather,
they vouchsafe the traditional character of the ‘Doric’ forms ὑπάτα, μέσσα, τρίτα, and
νεάτα in Philolaus fr. 6b DK.
58
I owe this observation to R. Janko (communication, 1998); for the phonology,
see Ruijgh (1961), 204–6; cf. Franklin (2006) 55 n. 42; Franklin (2011) 533. Although
ἁρμονία is not certainly attested in the sense ‘tuning’ before Lasus of Hermione (fr. 1
PMG 702), it probably does appear a century earlier in Sapph. fr. 70.9–11 ( ]αρμονίας
δ̣[ | ]αθην χόρον, ἄα[ | ]δ̣ε λίγηα.[ , where χόρον and λίγηα support a ‘lyric’ context).
Also relevant are early musical applications of the poetic ἀραρίσκω, with contexts
including lyre-construction (Hom.h.Merc. 50 ἤραρεν, with Franklin 2002a: 9) and
‘joining together’ song (Hom.h.Ap. 164: οὕτω σφιν καλὴ συνάρηρεν ἀοιδή, of the
Delian chorus), obviously implying careful achievement of purposeful pitches.
32 John C. Franklin
more general historical nature. In a paper for the Oxford Dithyramb
conference in 2004 I re-assessed the chronological evidence for the
practice of modulation (πολυφωνία, καμπαί, etc.), often associated
with the ‘New Music’ of the later fifth-century and contrasted in our
sources with a more ancient heptatonic artform.59 I concluded, from
several poetic passages and historiographical notices, that the basic
techniques could be traced back as early as the sixth or late seventh
century (Sakadas/Klonas), with the New Musicians representing a
tertiary stage of evolution (following the intermediate innovations of
Lasus, Simonides, and Pindar). Conversely, the practice of adhering
to a single ἁρμονία had special links to early lyre practice according to
a passage of ps.-Plutarch that must derive from Hellanicus’ work on
the Lesbian γένος of lyrists (including Terpander).60 This ancient
assertion, which one might otherwise suspect of being an historio-
graphical construct, is strikingly corroborated for the Bronze Age
Near East by the Hurrian hymns from Ugarit and the Middle Assyr-
ian song catalogue VAT 10101 (see p. 23 n.23 and p. 36). The Bronze
Age inheritance of the Lesbian γένος is nicely encapsulated in myths
that the lyre and/or head of Orpheus wafted to Lesbos after his
death.61
59
Franklin (2013).
60
[Plut.] De mus. 1133b (cf. 1137a–b): ‘In general, the style of citharody practiced
by Terpander persisted even unto the time of Phrynis as one which was altogether
simple. For in the old days it was not allowed to make citharodic compositions like
today, nor to transfer the ἁρμονίαι and the rhythms (sc. beyond their proper bound-
aries). For in the νόμοι they guarded the proper tuning for each.’ For a defence of this
tradition, see further Franklin (2002d) 698–9; Franklin (2012) 743–8 and passim with
Franklin (2013) 217–18.
61
For these traditions generally, see Pfister (1909–12) 213 n. 213; Power (2010)
390–1. I suspect, from a rough reconstruction of Hellanicus’ Karneian Singers
(Franklin 2012, esp. 29–30), that the historiographical ‘master-myth’ of lyric history
in [Nicom.] Exc. 1—beginning with Orpheus and ending with Terpander—is an
epitome of the ‘Bronze Age’ section of Hellanicus’ work. Note especially the embed-
ded hexametric fragment about Amphion—Ἀμφίωνα τὸν Θηβαῖον, ὃς ἐπὶ τῶν ἑπτὰ
χόρδων ἑπταπύλους τὰς Θήβας ᾠκοδόμησεν (cf. Franklin 2006: 56 n. 46), recalling that
there existed a verse-edition of Hellanicus’ history (FGrH 4 F 85a = Athen. 635e).
A key piece of the puzzle that has been neglected here is Bryennius Harmonica 1.1.54.
Hellanicus’ history has apparently also left an imprint on writers of the generation
following its publication; note especially the sphragis of Timotheus’ Persae, and the
flurry of interest in the expression ‘Asiatic kithara’ (Euripides, Aristophanes, et al.).
I gave a paper on this called ‘Euripides and the Archeology of Music’, at the
conference Music in Greek Drama: History, Theory and Practice, May 28–9, 2011,
University of California, Santa Cruz, USA. See also Franklin (2010) ‘Hellanicus
of Lesbos’.

These points suggest a re-interpretation of the Greco-Lydian
musical movement already discussed. Rather than taking Terpander
as symbolizing a seventh-century re-importation of the Mesopota-
mian musical knowledge as documented by the Bronze Age tablets,
I would now posit secondary Greco-Lydian exposure to contemporary
Mesopotamian and specifically Assyrian practice (itself probably a
cosmopolitan conglomerate). The sudden prominence in the Archaic
poets of harps (typically polychordal and certainly deriving from the
East),62 the two sizes of αὐλοί used in Alyattes’ military band and
apparently shown in a Neo-Assyrian relief,63 and the cultural prestige
of the Lydian court in the seventh and sixth centuries—these points
all suggest that the modulatory techniques which began to emerge at
just this time in Greece (the dates of Klonas and Sakadas match the
Lydian ascendancy) were at least partially stimulated by wider-
ranging musical fashions of the day. Importantly Terpander himself
is implicated in traditions of ‘polyphonic’ growth beyond the ancient
heptachord, including the same alleged scandal at Sparta otherwise
suffered by the New Musicians Phrynis and Timotheus.64
But what then becomes of the idea that Terpander invented the
seven-stringed lyre, which after all rests upon the traditional verses
attributed to him? Iconographic record aside, the fragment’s ‘four-
voiced song’, with its clear antithesis to ‘heptatonic φόρμιγξ’, remains
unexplained. And the sheer number of three- and four-stringed lyres
in Geometric art, combined with abundant ethnographic analogies
for epic ‘melodization’ using only a few pitches,65 does rather strongly
suggest that heptatonic lyres were not universal in the Dark Age.66
My current working hypothesis is that praise-singers of the
62
Franklin 2008, 201 n. 5.
63
BM 124922: Rashid 1984, 126 and fig. 145, with Cheng 2001, 35, who also notes
that ‘long pipes’ may be distinguished in Akkadian texts.
64
Terpander: Ps.-Arist. Prob. 19.32 (on ‘Dorian νήτη’, related to other Problems on
the transition from seven to eight strings: 19.7, 25, 44, 47); Plut. Inst. Lac.17.238c–d;
Ps.-Plut. De mus. 1140f. Certainly relevant is Timoth. Pers. fr. 15.224–5 (PMG 791),
although the point is obscured by textual and interpretative uncertainty; for the issue,
see Wilamowitz-Möllendorff (1903) 27, 68–9; Aron (1920) 33–4; Janssen (1984)
153ff.; Gostoli (1990) test. 46 and 113–14; Hordern (2002) 242–3; Ercoles (2010)
122–4; Power (2010) 338 n. 58. Phrynis: Plut. De prof. virt. 84a. Timotheus: Plut. Inst.
Lac.17.238c–d. Athenaeus also refers to the three times the Spartans rescued music:
Ath. 628b. For the traditions about Sparta, see further Power (2010) 172–3, 340–1, 536
and n. 347 with earlier bibliography.
65
See e.g. papers in Reichl (2000); cf. Franklin (2004).
66
So rightly West (1992b) 52.
34 John C. Franklin
Mycenaean period elaborated inherited diction and metrical forms
into extended heroic narratives, using a formulaic melodic language
(and perhaps already lyre-accompaniment) of limited tonal scope.67
Simultaneously a local version of the Mesopotamian diatonic tuning
system was cultivated primarily in cultic contexts, in emulation of
palatial practice in the Near East (with the cosmopolitan, syncretic
hymns from Ugarit a vital analogy). In the Iron Age these parallel arts
died or developed locally—both may have flourished on Lesbos—and
were eventually ‘recombined’ in the pan-Hellenic trends of the
Archaic period.68 Thus we may explain the tradition that Terpander
dressed the ‘ἔπη of Homer’ in the ‘μέλη of Orpheus’,69 an idea that
seems implicit in the Terpandrian fragment itself, where the poet’s
φόρμιγξ—a word with strongly epic connotations, versus the equally
ancient λύρα70—is startlingly described as ἑπτάτονος, a technical
epithet with clear melic overtones and probably meaning precisely
‘diatonic’ at this time.71 This theory of regionalism followed by
reconvergence—which is after all a well-known dimension of the
general historical sequence—would allow the ancient, inherited hep-
tatony of the Lesbian γένος to appear as a genuine novelty in other
parts of Greece, and so justify the traditional interpretation of the
verses credited to Terpander. After all, his reported professional
activity, outside the immediate Lesbo-Lydian environment, transpires
at Delphi and Sparta, which would place him before Dorian and
Central Greek audiences—that is, at the interface of Bronze Age
inheritance and the shifted demographics of the Early Iron Age.72
This proposed syncretism of distinct epic and melic traditions in the
seventh century (if not earlier) could also explain why the Archaic
Greek heptachord incorporates names derived from fingers (λιχανός,
μέση, παραμέση). West suggested that these might be a vestige of an
67
Indo-European dimension of Greek epic melodization: West (1981); Franklin
(2004).
68
For these points see Franklin (2011).
69
Alex. Polyhist. FGrH 273 F 77 (Ps.-Plut. De mus. 1132e–f), deriving from
Heraclid. Pont. fr. 157 (Ps.-Plut. De mus. 1132c); cf. Clem. Alex. Strom. 1.16.78;
Suda s.v. Τέρπανδρος. Cf. Franklin (2002c) 445–6 (allowing for the modifications to
the historical scenario proposed above); Franklin (2004) 244–5.
70
Λύρα is now indirectly attested at Mycenaean Thebes (TH Av 106.7): see with
references Franklin (2015) 433–44 and n. 62.
71
Franklin (2002d) 673–5.
72
The generations of the Lesbian γένος that followed Terpander were also remem-
bered especially for their performances at the Spartan Karneia: see Franklin (2012).

earlier nomenclature appropriate to an instrument of only three or
four strings, each named from a finger.73 He made this observation
without reference to the overall epicentricity of the Greek heptachord,
an issue with which he was not then familiar.74 But this distinctive
structure—an epicentric heptachord built around a nucleus of finger-
names—might well reflect the fusion of epic and melic lyre traditions
proposed here.75
MUSIC, TEXT, AND EPICENTRIC TONALITY
This section offers preliminary observations on a pair of complemen-

tary questions. What was the impact of epicentric tonality on poetic
‘text’, i.e. its musical form and realization in performance? And what
traces has the epicentric heptachord, as a cultural artifact, left in
Greek literature and the intellectual record more generally?
First there is the issue, for those who accept that Greek epicentric
tonality does indeed derive in some form from the Near East, of ‘what
else’ came along with it. Might we not have here a vital clue for
understanding how various thematic elements of early Greek litera-
ture find Near Eastern counterparts, since in both spheres ‘poetry’
was often musical? True, some classicists’ response to Martin West’s
catalogue in the massive East Face of Helicon has been that, after all,
‘there’s not that much there’; the proposed correspondences are
usually not terribly specific, so that independent development may
appear the simpler explanation.76 For my part, such elusiveness is just
73
West (1981) (noting only λιχανός and μέση). The historiographical tradition
regarding the transition from seven to eight strings (see n. 64 above) is often
concerned with the distinction between παραμέση and τρίτη and the history of
these terms’ usage, and that of conjunct and disjunct tetrachords constructed in relation
to μέση. I suspect that παράμεσος and τρίτος were interchangeable designations for the
same finger: see e.g. Galen De differentia pulsuum 8.544, lines 10–22.
74
He did not discuss the Mesopotamian musical texts until West (1993–4), when
he dismissed the parallel epicentric arrangements as likely to be coincidence; but when
presented with the further parallels discussed above, he wrote ‘Your main thesis, that
“heptadiatonic” music came to Greece (indirectly) from Mesopotamia, I find quite
plausible. This music is after all very much bound up with the stringed instruments
that came from that direction’ (communication, 28 April, 2000).
75
See for now ch. 9 of Franklin (2002b).
76
West (1997). Quotation: anonymous (communication, 2006).
36 John C. Franklin
what I would expect from the historical circumstances. The surviving
texts are but the merest fraction of what was once composed in a
multidirectional game of Chinese whispers (if one may use the
expression) spanning huge distances and many centuries. And each
participating culture would inevitably leave a distinctive imprint on
mythemes and other motifs of otherwise ‘international’ circulation.
Finally, if one places the most significant phase of ‘contact’ in the Late
Bronze Age, as I would—since the period’s palace networks promoted
tightly focused elite cultural exchanges across large distances77—this
allows for many centuries of subsequent, independent development
on the Greek side, especially across the Early Iron Age and into the
time for which we first have texts. I certainly do not wish to exagger-
ate the cultural links; I am in fact a minimalist, preferring the least
intensive contact that will suffice to explain the data. It might be, for
instance, that epicentric tonality was adopted by Aegean lyre-singers
as being a maximally efficient—and available—conceptual technol-
ogy for the varied musical application of heptatonic/diatonic tonal
material, a scenario that is conceivable without requiring any ‘higher’,
literary influence.
Still, the generic contexts for which the Mesopotamian tunings are
attested offer striking correspondences with Archaic Greek lyric
practice. The Middle Assyrian song-catalogue VAT 10101 (see
p. 32) organizes a singer’s repertoire of some 360 titles into thirty-
two categories/genres of song, of which only two employed the
heptatonic/diatonic tunings. Evidently the heptatonic/diatonic cycle
had quite specialized application in the Near East, and we must not let
the universality of heptatonic/diatonic scales in our own music mis-
lead us into conflating these specific tonal constructions with ‘music’
generally in antiquity.78 The songs called šitru are connected to
choral, ensemble performance; here only two of the seven tunings
appear (embūbu and pītu).79 That the ‘breast songs’ (irtu) are per-
sonal love songs is indicated by the titles themselves; these use all
seven of the tunings (some more than others).80 Finally, the hymns
from Ugarit show that the tunings could also be used for divine
77
For the musical dimensions of this environment, see as a whole Franklin (2015).
78
Even our ‘atonal’ music has an inescapably heptatonic/diatonic basis, since this
provides the twelve tones to which it is limited.
79
For šitru(m), see now Ziegler (2007) 11, 13, and n. 35 with further references
(superseding Kilmer 1971: 147; Kilmer 1997: 475; etc.).
80
For irtu, see Held (1961) with Kilmer (1971) 138 n. 24; Kilmer (1997) 475.

hymns, at least in Syro-Hurrian lyric tradition.81 Choral song, love-
lyric, divine hymns—these are the great types of Greek lyre μέλος.
Moreover, the relevant sections of VAT 10101 and the Hurrian
hymns show that it was conventional for a song to remain within a
single tuning,82 something that we are told was basic to early Greek
lyric.83 This practice has important implications for the pre-history of
musical ‘ethos’—that topic of primary concern for Greek musical
thinkers of the Classical period84—since clearly each tuning (and
any mode derived therefrom85) must have had its own distinctive
character. The foregoing collection of correspondences in genre and
harmonic practice strikes me, once again, as too specific to be pure
coincidence; together they provide good circumstantial support for
the hypothesis of epicentric tonality as an early, supralocal approach
to lyre-playing.
But even if one rejects an historical connection with the Near East
altogether, it remains certain that heptatonic/diatonic tunings and an
epicentric-tonal approach to their practical implementation were
important elements of Archaic Greek lyric. These are welcome and
rather concrete facts for any attempt to reimagine the musical dimen-
sions of poetic texts, or ‘reconstruct’ performances, since they con-
tribute strong new formal constraints—perhaps not permitting such
specific recoveries as metre and rhythm, but still far better defined
than passing internal references to ‘Phrygian μέλος’ and the like.86
True, later historiographical allegations about e.g. the modulatory
scheme of the τριμελὴς νόμος, or Sappho’s use of the Mixolydian,
do offer (potentially) precise harmonic information in a very few
81
For the distinctive identity of Syro-Hurrian lyric (and its use of the kinnāru), see
Franklin (2015) 96–104.
82
Even if we suppose the occasional neighbouring modulation (as suggested by
Hagel 2005b), a song remained harmonically identifiable as some specific tuning.
83
See n. 60.
84
See inter al. Anderson (1966); Rossi (1988); Thorp (1991); Pagliara (2000);
Lynch (2013); Lynch (2016).
85
I ought perhaps to clarify the difference between ‘tuning’ and ‘mode’ as I use
them here. In Indian tradition, modes (raga) are distinguished from underlying
tunings; the former are elaborations of the latter, which naturally has an influence
on their development. I thus use ‘mode’ in a broader sense than its old application to
octave ‘species’, which I here designate by ‘tuning’. But neither term is quite exact for
epicentric tonality, as I understand it, since this system both provided the ‘species’ and
influenced the elaboration of mode (via its practical focus on μέση).
86
Alcm. 126 PMGF; Stesich. 212 PMGF.
38 John C. Franklin
specific cases.87 But if we may trust the sources, epicentric tonality was
a quite general feature of early lyric. Alongside this we must probably
set the regular custom of melodizing the pitch contour of poetic
diction, a fairly consistent practice in the extant musical scores—
especially those of a simpler, more traditional character—and already
detectable, in some kindred form, in the accent patterns of epic.88
(Whether we are to assume pre-composed and fixed melodies for
Archaic lyric, or something more along the lines of composition-in-
performance within fixed formal constraints, is up for debate.)89
When epicentric tonality and ‘accent-singing’ are taken together,
the possibilities for realizing Archaic melody become much more
sharply defined. Μέση was like the sun around which the other strings
revolved, giving to each its tonal force—what Aristoxenus would later
call δύναμις90—and steadily reinforcing these relations through its
frequent repetition. Admittedly we must still resort to guesswork
about what exact tunings were favoured, and the kinds of traditional
modes (νόμοι) that were developed from these. The most obvious
candidates are the conjunct and disjunct ‘Dorian’ structures that are
central to later theory, allowing for variations induced by devout
insistence on seven strings.91 By way of example Figure 1.4 gives a
87
Τριμελὴς νόμος: Ps.-Plut. De mus. 1134a–b; cf. 1131f–1132a (= FGrH 550); cf.
Franklin (2013) 223–4. Sappho/Mixolydian: Aristox. fr. 81 = Ps.-Plut. De mus. 1136.
88
For the observable ‘rules’ governing melodization of accent, see West (1992b)
199–200, noting that the correspondence is especially strict in the Delphic paeans,
with their archaizing tonal material. For Homeric accent patterns as a kind of melodic
residue, see Hagel (1994).
89
See further D’Angour (this volume).
90
A key passage here is Harm. 47: ‘Why is it that there is one interval between
μέση and παραμέση and again between μέση and both ὑπάτη and as many others as do
not change pitch, while it must be ruled that there are many intervals between μέση
and λιχανός?’ Here Aristoxenus betrays his knowledge of the old epicentric
perspective—familiar to the lesser colleagues whose musings are reflected in the
Aristotelian Problems—by imagining all possible intervals that might be taken with
μέση, dividing them into those of fixed and variable size, as though this is how the
problem would naturally be—or was in fact—posed (cf. διὰ τί, the familiar formula of
the Problems). The connection between μέση and δύναμις is stated more clearly by
Cleonides in his ‘Aristoxenian handbook’ (11, MSG 202.3–5): ἀπὸ δὲ τῆς μέσης καὶ
τῶν λοιπῶν φθόγγων αἱ δυνάμεις γνωρίζονται, τὸ γὰρ πῶς ἔχειν ἕκαστον αὐτῶν πρὸς
τὴν μέσην φανερῶς γίνεται (‘And the functions of the rest of the notes are known from
μέση, for how each of them is clearly arises in relation to μέση’). See further Franklin
(2002b) 274–7 (§10.29–33).
91
Devotion to seven: see n. 36. In their textbook, form these are respectively
E-F-G-A-Bb-c-d and E-F-G-A-B-c-d-e. With only seven strings the disjunct ‘hepta-
chord’ often included ‘Dorian νήτη’, thus creating a ‘gap’ of a third (E-F-G-A-B-c-e,
Figure 1.4. ‘Recomposition’ of Sappho 1, illustrating epicentric tonality and

accent-melody in conjunct heptachord E-F-G-A-Bb-c-d.
recomposition of Sappho 1, adhering to epicentric tonality, accent

contour, and the textbook conjunct heptachord.
But one can hardly doubt that much greater variety existed. Hep-
tatonic/diatonic tuning in itself imposes well-defined parameters, and
while we need not insist that all Archaic ἁρμονίαι were of this type, it
would be musically naive to deny that the melic poets could distribute
tones and semitones wherever it suited their purposes via ἡ λῆψις διὰ
συμφωνίας (along with any microtonal shadings that may have been
fashionable).92 There will thus have been some de facto overlap with
E-F-G-A-B-d-e). (The modern values E-F-G etc. are a conventional means of illus-
trating relative pitch in ancient tunings, with lower case denoting a further, higher
octave; equal temperament is not implied). Key sources here are Philolaus fr. 6b and
several Aristotelian Problems (see n. 61). But we should be cautious about overgen-
eralizing from this material (Philolaus himself was clearly aware of the tone/semitone
grid behind his ‘gap’: see Franklin 2002b, ch. 8).
92
For tuning by consonance as a usual preliminary to microtonal shadings (χρόαι),
see generally Franklin (2005).
40 John C. Franklin
the seven diatonic tunings of the Mesopotamian cycle, whether or not
Greek lyrists of the Archaic period (still?) understood the progression
from one to the next in precisely the way that is documented by the
Retuning Text.93
Until now I have left the surviving musical scores out of consider-
ation. The relevance of these much-later documents to the Archaic
epicentric tonality alluded to by our sources is complicated by the
facts of Greek musical development. The original epicentric structure
was first compromised in the fifth century when additional strings
began to be added. Several ancient lists of ‘string-adders’ survive, but
these are artificial constructions of very limited direct value.94 Much
more important is the ancient historiographical debate about the
original distinction between παραμέση and τρίτη; this must relate to
the addition of a standard eighth string by c.480–460, to judge from
ceramic evidence and a tradition that Simonides added the eighth
string or τὸν τρίτον φθόγγον.95 Although Ion of Chios sang of his
remarkable eleven-stringed instrument, and lyres with ten–twelve
strings are indeed sometimes shown in fifth- and fourth-century vase
painting,96 lyre-singers seem to have settled on a new standard of nine
strings. This has not been well-appreciated until Hagel’s recent work;
we have been distracted by the two-octave, rather αὐλός-driven Perfect
System that Aristoxenian theory used as a reference structure in its set
of thirteen (later fifteen) keys (τόνοι); for this used only eight unique
string names (the ancient seven, plus τρίτη) in its central (historically
‘lyric’) octave around μέση, qualified as needed for referring to higher
and lower tetrachords (e.g. ὑπάτη ὑπατῶν).97 But several sources do
attest one further string called either ὑπερυπάτη or διάπεμπτος.98 Both
93
If the ‘transmission’ goes back to the Mycenaean period, considerable changes of
perspective may have developed in the intervening centuries. I thank Stefan Hagel for
useful comments on this point (Corfu, 7/2007).
94
These are surveyed by Hagel (2009) 80–7.
95
Pliny Nat. 7.204; Suda s.v. Σιμωνίδης. Cf. Franklin (2002d) 686 n. 45.
96
For various thoughts on Ion of Chios fr. 32 West, see Levin (1961); Comotti
(1972); West (1992a) 25–6; West (1992b) 227, 357; Hagel (2000) 52–3; Franklin
(2002d) 687–8, 693–4; Power (2007); Hagel (2009) 377–8. Ceramic evidence: West
(1992b) 63.
97
The τόνοι are a series of self-contained Perfect Systems whose μέσαι were
staggered at semitone intervals. For the development of this system, see now Hagel
(2009), ch. 1.
98
West (1992b) 224 n. 14, tentatively took Πράτα (‘first’) in the Argive inscription
SEG 30.382 (see n. 57) as a further alternative for ὑπερυπάτη/διάπεμπτος. If this is
right, one must assume the superimposition of a non-epicentric perspective, like the

names are clearly accretions to the early heptachord’s epicentric
nomenclature, and yet are only meaningful in relation thereto.
Whereas the superlatives ὑπάτη (‘topmost’) and νήτη (‘last’) clearly
marked original physical limits on the lyre’s crossbar,99 ὑπερυπάτη
(‘above the topmost’) still assumes that one’s orientation is outwards
from μέση in the way that Aristotle treated as normal.100 Similarly,
διάπεμπτος shows that this string’s usual (perhaps exclusive) function
was to form a 3:2 fifth below μέση. This created an overall span of a
ninth, so that the ancient ‘Lydian’ (once known as ‘Dorian’)101 octave
E-F-G-A-B-c-d-e—central to Greek music theory—was given a lower
extension to D by practising lyre-singers. The prize document here,
and our source for the term διάπεμπτος, is the so-called Koine Hor-
masia (roughly ‘Common Procedure’), which Hagel has recently
shown to be a fragment of an instruction manual for tuning a nine-
stringed κιθάρα through several diatonic tunings, of which only the
initial Lydian and its modification to the adjacent Hypolydian survive.
This remarkable document vitally attests the survival of epicentric
tonality down into the Roman period, for the tuning procedure begins
by establishing μέση with further intervals taken διὰ συμφωνίας
throughout.102 It also lets us integrate the Mesomedes hymns and
the Seikilos song into our history of epicentric tonality. For while
these pieces are ideally suited to a lyre of nine strings tuned precisely
as in the Koine Hormasia,103 yet none employs μέση (‘A’) as a tonal
center in the way described by earlier sources. This function is served
rather by ὑπάτη (‘E) in the Invocation of the Muse, Invocation of
linear presentation of octaves species or the Perfect System in the musicographers.

The third-century BC date makes this possible in an age of theory and musical literacy;
but why then do the inscriptions make a point of including the epicentric boundaries
Νήτα, Μέσσα, and Ὑπάτα? Hagel (2009) 27 n. 80 and 287 n. 68 suggests taking πράτας
rather as a gloss or description of the adjacent ὑπάτας, with reference to this
string’s function as an independent, non-epicentric tonal centre (paired at a fifth
with παραμέση) in three of the Mesomedes pieces (see further p. 42).
99
West (1992b) 221 with sources in n. 9.
100
See n. 31.
101
As Hagel has shown, the (ancient) characterization of this octave as ‘Lydian’
rather than the expected (and more ancient) ‘Dorian’ derives from several stages of
development in the notation/key system: see ch. 1 of Hagel (2009).
102
Hagel (2009) 122–34. Μέση itself is taken from προσλαμβανόμενος, an octave
lower; but as Hagel rightly argues this must be a reference tone given by e.g. αὐλός,
voice, or some other instrument which the lyre needed to accommodate.
103
The vocal melody occasionally exceeds this range, but octave counterparts exist
within the central compass of the instrument.
42 John C. Franklin
Calliope and Apollo, and Hymn to the Sun (DAGM 24–7); or by λιχανός
(‘G’) in Seikilos (DAGM 23), the Hymn to Nemesis (DAGM 28), and as
a secondary tonality in the Hymn to the Sun. As Hagel rightly argues, in
reference not only to these pieces but also the evidence of Ptolemy,
such ‘displacements’ are related to the permanent addition of
ὑπερυπάτη/διάπεμπτος, which presented a new tonal axis of D-A-d
alongside the old E-A-e; this helped accommodate modulations to
other modes with less retuning than in former times, and this in turn
probably encouraged the emergence of new modal norms.104 Never-
theless, the Koine Hormasia documents clearly that μέση ‘never ceased
being acknowledged as the “leader”’.105 And this is corroborated by
Mesomedes himself, whose name should mean something like ‘coun-
seled by Μέση’ (compare e.g. Diomedes)106—a remarkable testimony
for understanding the lyre-singers’ relationship with their instruments
(see p. 44–5).
The long-enduring importance of μέση for establishing tonal
values—and hence of the epicentric perspective itself—is further
seen in the various microtonal measurements that theorists offered
over the centuries for the basic Lydian/Dorian octave and its three γένη
(diatonic, chromatic, enharmonic). These figures, which the sources
express as ratios between successive degrees—many rather abstruse—
have often been supposed to owe more to mathematical tidiness and/
or fantasy than musical reality. But an analysis of cross-relationships—
that is, when the given figures are used to calculate ratios between non-
successive scale degrees—reveals that in many cases non-adjacent
notes are related to each other by much less abstruse ratios like the
audibly resonant 5:4 and 6:5 thirds. Moreover, in many cases these
intervals are constructed in relation to μέση. The two most striking
examples are the enharmonic of Archytas, our oldest such authority
(early fourth century BC); and the chromatic of Didymus (first century
107
AD, Figure 1.5). It may still be that some of these specific numbers
are indeed products of theoretical ingenuity.108 Even so, the general
emphasis on μέση shows that such mathematical schemes were still
responding to an important musical reality.
104 105
Cf. Hagel (2009), ch. 4. Hagel (2009) 134.
106
I owe this observation to S. Hagel, communication June 2013.
107
For the various other schemes not illustrated here, see the diagrams in Franklin
(2005), building upon Barker (1984–9) 2.46–52.
108
Cf. Hagel (2006).
h ph l m pm t pn n
h 1 0.9373 0.8997 3:4 2:3 0.6249 0.5998 1:2
2) 5:4
ph 1.0668 1 0.9598 4:5 0.7112 2:3 0.6399 0.5334 6:5
10:9
l 10:9 1.0419 1 5:6 0.741 0.6946 2:3 0.556 3) 3:2
6:5
m 4:3 5:4 6:5 1 8:9 5:6 4:5 2:3 5:4
10:9
pm 3:2 1.4061 1.3496 9:8 1 0.9373 0.8997 3:4 4) 5:4
6:5
5) 3:2
t 1.6003 3:2 1.4398 6:5 1.0668 1 0.9598 4:5 5:4
6:5
pn 1.6673 1.5628 3:2 5:4 10:9 1.0419 1 5:6
n 2:1 1.8746 1.7994 3:2 4:3 5:4 6:5 1
Figure 1.5. Didymus’ chromatic γένος, expressed in matrix showing ratios

between all string pairs, with decimal figures replaced by resonant/epimoric
ratios (3:2, 4:3, 5:4, etc.) where applicable.
From around the same time as Didymus come two formulations of

the Harmony of the Spheres deriving from the multiform Pythagor-
ean tradition, one in Nicomachus’ Encheiridion, another in the
Excerpts attributed to him. In both, the seven strings, each assigned
to a heavenly body or celestial sphere, are listed in sequences that,
though slightly different from one another, alike respect the epi-
centric order—note especially their agreement on μέση as the Sun—
thus showing that this was the essential organizing principle:
Nicom. Ench. 3: [Nicom.] Exc. 3109
2. νήτη/Selene-Moon 2. νήτη/Selene-Moon
4. παρανήτη/Aphrodite 5. παρανήτη/Hermes(?)
7. παραμέση/Hermes 7. τρίτη/Aphrodite(?)
5. μέση/Helios-Sun 3. μέση/Helios-Sun
6. λιχανός/ὑπερμέση110/Ares 6. ὑπερμέση/Ares
3. παρυπάτη/Zeus 4. παρυπάτη/Zeus
1. ὑπάτη/Kronos 1. ὑπάτη/Fixed Stars/Kronos
109
This passage presents several uncertainties and variants acknowledged by the
text itself. My question marks next to παρανήτη and τρίτη reflect the author’s correct
hypothesis of scribal error, although it is not immediately clear which god should go
with each, since Nicom. Ench. 3 cannot be used as a parallel without circular argument
(τὴν δὲ παρανήτην οὐ κατὰ τὸν Ἑρμῆν, ἀλλὰ κατὰ τὴν Ἀφροδίτην, ἀτάκτως, εἰ μὴ
γραφικὸν εἴη τὸ πταῖσμα . . . τὴν δὲ τρίτην κατὰ τὴν Ἀφροδίτην). He also offers an
alternative association for ὑπάτη, either Kronos or ἀπὸ τῆς ἀπλανοῦς (sc. σφαῖρα), and
mentions a version with strings/planets in opposite order. For further related material,
see Jan (1894); Zaminer (1984) 7–8, 21–5; Hagel (2005a).
110
Ὑπερμέση is an alternative designation in these texts for λιχανός, and was
presumably favoured for more clearly eliciting the epicentric structure; whether it
has an equally traditional basis is unclear.
44 John C. Franklin
These structures must be related to the report of Boethius, alluding to
the lost book of Nicomachus, that Terpander completed the hepta-
chord in imitation of the seven planets.111
With these remarkable cosmological visions we move from our
first question, the relationship between epicentric tonality and music/
text, to the second, its broader impact on the Greek intellectual—even
spiritual—imagination. Most striking of all, to my mind, is the treat-
ment of Μέση, Ὑπάτη, and Νήτη—the trio of boundary strings that is
a shorthand for the epicentric heptachord112—as Muses at Delphi
according to Plutarch in the second century AD (whence my title The
Middle Muse).113 At first glance this may seem a late and whimsical
artifice. Originally I took it to be an example of what in ethnomusic-
ology has been called ‘museum effect’—intentional preservation of
older musical forms/ideas within a special social space, theoretically
protected from the innovations of a musical mainstream.114 In our
example, Greek cultural memory of an essential technical aspect of
early lyric would have been enshrined, appropriately, at the sacred
site of Apollo (that Delphi was regarded as the centre of the world
might also be relevant). But the third-century BC Argive inscription
mentioned above115 shows that Plutarch’s string-Muses were no one-
off conceit at Delphi, but represent a broader set of conceptions with
much older roots in Greek lyric practice. The assignment of a divine
power to individual musical tones is of course also found in Plato’s
Myth of Er, where a separate Siren revolves (n.b.) upon each and
every tone of the cosmic ἁρμονία (doubtless diatonic as in the Ti-
maeus).116 But Muses are especially interesting for their implications
of music-cognition—of a reverent, contemplative relationship
between musician and instrument in which the lyre itself plays an
active, even dominant role.117 Recall that Apollo’s own reaction, in
the Homeric Hymn to Hermes, to his brother’s new, seven-stringed
111
Boeth. De inst. mus. 1.20 (206.10f.): septimus nervus a Terpandro Lesbio
secundum septem scilicet planetarum similitudinem.
112
See nn. 31 and 57.
113
Plut. Quaest. conviv. 744c, 745b; cf. Ps.-Censor. De mus. 6.610.1f.
114 115
For the term, Nettl (1985) 28. SEG 30.382: see n. 54.
116
Pl. Rep. 617b5–6: Σειρῆνα συμπεριφερομένην, φωνὴν μίαν ἱεῖσαν, ἕνα τόνον; cf.
Philostr. Imag. 1.10.15.
117
In the Hymn to Hermes, the singer’s subservience to lyre is also expressed in
erotic terms. It arouses in Apollo ἔρος ἀμήχανος, ‘inescapable love’ (434; cf. Hermes as
‘deviser’, μηχανιῶτα, 436), that melic trope for love as a cause of grief, with the lover
enthralled to the beloved. Hence at 447 τίς Μοῦσα ἀμηχανέων μελεδώνων; will be a

lyre was ‘Who is this Muse?’ (τίς Μοῦσα; 447). Hermes responds
(482–4) by characterizing ‘her’ (λύρα is also feminine) as a teacher:
ὅς τις ἂν αὐτὴν
τέχνῃ καὶ σοφίῃ δεδαημένος ἐξερεείνῃ
φθεγγομένη παντοῖα νόῳ χαρίεντα διδάσκει
Whoever, learned in skill and wisdom, enquires of her, she teaches,
uttering all sorts of things pleasing to the mind.118
The Delphic and Argive string-Muses must represent a similar con-
ception. But they preciously specify what exactly gave the seven-
stringed lyre its ‘teaching’ properties. The epicentric Muses lead the
lyrist—recall that μέση was sometimes called ἡγεμών119—into a
treasure house of ideas and inspiration, provided they be cultivated
with the devotions of ‘skill’ and ‘wisdom’. The name of Mesomedes
(see p. 42), and our new understanding of the Koine Hormasia, show
that the string-Muses were more than a Delphic museum-piece in
Plutarch’s day—though perhaps one may still detect an antiquarian
flavour in their limitation to the three boundaries of the old hepta-
chord in an age of nine-stringed instruments and modulation-keys.120
Thus we find that epicentric tonality, besides its welcome contri-
bution to our understanding of tonal and melodic phenomena, adds
considerable down-to-earth texture for our attempts to grapple with
the relationship between singer and song. Familiar translations of
ἁρμονία as a ‘pattern of notes’, ‘attunement’, and so on are accurate so
far as they go, but promote an overly abstract understanding of tonal
‘material’ that ellipses the complex and intimate intercourse between
lyrist and his or her tangible interface with ‘god-spoken song’ (θέσπις
ἀοιδή). The early profundity of such conceptions is best seen in the
practice of divinizing instruments that is richly attested for the
double-entendre, ‘Who is this Muse who causes inescapable cares’—inverting the

traditional motif that music can cure cares (Hes. Th. 98–103; Hom.h.Ap. 182–93; both
passages envision choral song to lyre-music). The theme reaches its climax (and
resolution) when Hermes characterizes the lyre as a clear-voiced courtesan (ἑταίρη)
whom Apollo can hold in his arms and engage with expertly (εὐμόλπει μετὰ χερσὶν
ἔχων λιγύφωνον ἑταίρην / καλὰ καὶ εὖ κατὰ κόσμον ἐπιστάμενος ἀγορεύειν, 478–9).
118
Compare the description of Linus in Hesiod fr. 306 M-W (PEG T 3): παντοίης
σοφίης δεδαηκότα. For these passages, see further Franklin (2006) 61–2; Franklin
(2015) 6 and n. 32, 307 and n. 159.
119
[Arist.] Prob. 19.33; [Plut.] De mus. 1135a.
120
Recall that Πράτα in the Argive inscription might classify ὑπερυπάτη as a fourth
Muse: see n. 98.
46 John C. Franklin
Bronze Age Near East (but with close parallels in some African
traditions).121 The LM III Chania pyxis with its musician, oversized
lyre, and bird-epiphany clearly expresses an equally awesome
encounter with musical divinity.122 Would such ideas have seemed
archaic in sixth- and fifth-century Greece? I would not underestimate
the religious intensity of Pindar’s opening invocation, in Pythian 1, of
Apollo’s ‘golden phorminx’, characterized by the poet as the true
leader of his choral song and dance.123 And by now we should be
well justified in supposing that Pindar, like other early Greek poets
who must have invoked their instruments in this way,124 was ever-
conscious of the epicentric heptachord as a medium through which
his musical mind would attain the song for which he prayed.125
121 122
Franklin (2015) 6, 22–3, 231–5, 309–10, and passim. See n. 51.
123
Pind. Pyth. 1.1–4: Χρυσέα φόρμιγξ, Ἀπόλλωνος καὶ ἰοπλοκάμων / σύνδικον
Μοισᾶν κτέανον· τᾶς ἀκούει μὲν βάσις ἀγλαΐας ἀρχά, / πείθονται δ’ ἀοιδοὶ σάμασιν /
ἁγησιχόρων ὁπόταν προοιμίων ἀμβολὰς τεύχῃς ἐλελιζομένα (‘Golden phorminx—
Apollo’s and the dark-tressed / Muses’ joint possession—whom the dance-step heeds,
the beginning of festivity, / And singers obey your signs / When thrumming you fashion
beginnings of chorus-leading preludes’). For this apostrophe as a late echo of ancient
ideas about the (divine) voice of (divine) instruments, see Franklin (2015) 235.
124
Compare Psalm 57:8–9 and 108:1–3, with Franklin (2015) 163–4.
125
We should stay alert for further literary allusions. Zaminer (1984) assembled an
array of promising possibilities in his study of the ὑπάτη-μέση-νήτη trope that call for
re-examination. The following may also be noted as potentially relevant: Theog. 1–4;
Heraclit. fr. 10 DK ([Arist.] De mundo 5.396b7); Philolaus fr. 6a and 6b, the connec-
tion between which has seemed obscure (note that Zaminer 1984: 25 treats fr. 17 as
relevant to μέση); Scythinus fr. 1 West (Plut. De Pyth. orac. 402a; cf. Burkert 1972: 320
and n. 107, 335–6); there are probably also connections with ideas of ‘middleness’, the
μεσότης of Plato and Aristotle (cf. Franklin 2002b: 253 §9.37).
The Musical Setting of Ancient

Greek Texts
Armand D’Angour
Scholarly examination of the few surviving ancient Greek musical

documents has allowed musicians to form a reasonable sense of the
way the rhythms and melodies were intended to sound in practice.
However, while many investigators have drawn on the work of
ancient theorists to explain the musical features evident in the
documents, few have attempted to consider specific ways in which
the music itself—that is, both the rhythms and melodies that can
be derived from the texts and markings in the documents—may be
related to the words that in most cases it accompanied or adorned.
In this chapter I propose to offer some suggestions about how the
musical expression attached to, or reconstructed to accompany, three
specific sets of texts may have been intended to contribute to their
significance and effect in practice, and to ask what conclusions might
emerge from them regarding the continuity or otherwise of ancient
practices of rhythmicization and melodization.
The texts in question comprise the opening lines of Homer’s Iliad
and Odyssey, with the accompanying melodic element derived from
the pitch-accents of the words; part of a chorus from Euripides’
Orestes with ancient musical notation preserved on papyrus; and
the ‘Seikilos song’ inscribed on a 2nd-century AD grave stele from
Asia Minor.1 These texts date from different periods spanning around
a thousand years, but they have in common the fact that all were
1
Orestes fragment, Seikilos stele: DAGM (no. 3) 12–17, (no. 23) 88–91.
48 Armand D’Angour
originally conceived, like the vast majority of poetry in archaic and
classical Greece, as sung music—words and music combined. They
will also have been subject to specific instrumental and performance
realizations of different kinds at different times, for which evidence is
solely circumstantial. In practice it is bound to have varied, depending
on the particular occasion of performance and the resources available
to performers.
We find virtually no comments in the work of ancient authors
regarding the aesthetic and phonic effects of a particular song or
passage of song such as these texts represent. While ancient discus-
sions abound regarding the ethos and effects of μουσική in general,
and of the ethical and, to a lesser degree, aesthetic effects of different
modes or instruments, one searches in vain for the description or
analysis in musical terms of any substantial poem or text. The passage
in which Dionysius of Halicarnassus, writing in the first century BC,
analyses the musical effects of some lines of Euripidean choral song is
uncommon enough for it to be accorded the status of a ‘document of
ancient music’ in its own right.2 But in the writings of earlier authors,
even those of musically-engaged thinkers such as Plato and Aristotle
and harmonic theorists such as Aristoxenus, there is no similar
commentary that might allow us to understand better the way the
melody or rhythm of, say, an ode of Pindar or a Sophoclean chorus
were heard in ancient times. Rare and passing mention is found in
classical writers about the musical effect of the works of composers
such as the tragedian Phrynichus of Athens, who ‘was always sipping
on the nectar of ambrosial melodies (μέλη) to bring forth sweet song
(φέρων γλυκεῖαν ᾠδάν)’;3 or such as Tynnichus of Chalcis ‘who never
composed a single poem that one would think worth mention other
than the paean which everyone sings, virtually the most beautiful of
all songs (μέλη), simply (as he says himself) an invention of
the Muses’.4 But nowhere do we find an articulation of the reasons
why particular μέλη should be honoured for such qualities as sweet-
ness or beauty, let alone a description of the specific musical features
that might be thought to bring about such responses.5
Having an understanding of how a particular song or piece of
music sounded is not the same as having a sense of how it was
heard by listeners in ancient times. For the latter purpose, the
2 3
DAGM 2, pages 10–11. Ar. Av. 748–51 (cf. Vesp. 220).
4 5
Pl. Ion 534d5–e1. Cf. D’Angour (2015) 189–92.
The Musical Setting of Ancient Greek Texts 49

comic parodies of the dramatists’ songs in Aristophanes’ Frogs offer
valuable if partial (and, given the context, unstraightforward) evi-
dence for the way the effects of melody or rhythm in specific instances
might have been received by contemporaries.6 Otherwise we are
largely dependent on authors of the Roman period and later for
scattered and unsystematic insights into the musical impact of
songs or poetic compositions. Thus Dionysius of Halicarnassus illus-
trates the way specific Homeric verses were felt to deploy rhythmical
effects;7 Pollux (2nd century AD) preserves details about the structure
and intended effects of the Pythikos nomos, a purely instrumental
piece for the aulos;8 and Aelian (early 3rd century) records how the
sixth-century statesman Solon of Athens, entranced by his nephew’s
performance of a μέλος of Sappho, asked to be taught it ‘so that I may
learn it and die’.9 In Aelian’s story neither is the poem in question
identified nor the precise basis of Solon’s enthusiasm. It is therefore
unclear how far the reported response should be thought to relate to
the words of the μέλος, rather than to its rhythmic or melodic
expression or to the particular vocal and instrumental virtues dis-
played on the occasion (though it is likely to have depended on a
combination of all these factors). Aelian’s account further exemplifies
how the attempt to find musical commentary on a particular text is
impeded by the tendency of ancient authors to conflate words and
music when commenting on the effect of μουσική, a term which in
classical times comprised both equally.10 By the time a critic such as
Longinus could present a descriptive interpretation of a specific song
by Sappho, the musical dimension is submerged, and his treatment of
a substantial portion of the poem deals exclusively with the way style
and imagery achieves sublimity.11
What accounts for ancient authors’ apparent lack of interest in
recording and preserving the specific melodies that formed such a
large part of their musico-literary heritage? The effects of melody,
while generally considered secondary to that of rhythm, were not a
6
Ar. Ran. 1264–95, 1309–63.
7
Dion. Hal. De comp. 20 (see Phillips in this volume with further references).
Aristides Quintilianus 69–75 (which may contain material drawn from Aristoxenus)
also analyses Homeric verses for rhythmical effects.
8 9
Poll. 4.84; cf. Strab. 9.3.10. Aelian ap. Stob. 3.29.58.
10
For an approach to the distinction of words and music in ancient discussions of
μουσική, see D’Angour (2015).
11
[Long.] De subl. 10.1–3.
negligible aspect of a song’s power in ancient ears.12 But while the
musical realization of a song might have made a considerable differ-
ence to its actual reception, the absence of comment on the nature of
a particular melodic line or passage confirms that only in rare cases
was the tune considered to be a dominant or even particularly
memorable feature.
In fact, ancient philosophers and musical theorists dwell far less
on μέλος than on ἁρμονία, the structure of pitches from which
an individual μέλος would have been derived.13 Classical poet-
composers and singer-performers, for lack of a system of vocal
notation (probably invented in the fifth century BC), will in most
cases have employed variable, orally transmitted melodic motives
conforming to appropriate ἁρμονίαι.14 As with oral folk music tradi-
tions universally, melody is likely to have been employed, for the most
part, in a flexible and relatively free fashion.15 Consequently, non-
notated melodies, including most sung texts until around the mid-
fifth century, would rarely have been felt to be determinate or to carry
authorial status. Rhythm, however, being a function of the syllabic
quantities of words, was at the author’s command; and during the
earlier period of Greek musical history it was considered of greater
importance than melody.16 But equally, given that the rhythms that
arose from words—iambic, dactylic, paeonic, and so on—rapidly
became conventional within their generic contexts, their musical
effects in a particular song or passage were apt to attract comment
only if, for instance, they were heard as unusual or wilfully uncon-
ventional (as in the case of the extended Euripidean melisma on
εἱλίσσετε parodied by Aristophanes).17
12
West (1992b) 129–30 cites comments on the importance of rhythm vis-à-vis
melody; but numerous passages of ancient poetry suggest a play on words between
μέλη, songs, and μέλει, ‘it matters’: D’Angour (2005) 99.
13
While prepared to discuss the ethos of different ἁρμονίαι at length, Plato was
suspicious and dismissive of the effects of μέλος; see Peponi in this volume.
14
For example, Aristotle (Pol.1342a32–b12) relates that Philoxenus was ‘forced’ to use
the Phrygian ἁρμονία when composing music for a dithyramb: D’Angour (2011) 207.
15
Nettl (2005) 113–15. This is an uncontroversial point for ethnomusicologists,
but it bears repetition, as the standard model for modern Western music is to privilege
‘the music’ in its own right.
16
Ps.-Plut. De mus. 1138bc. How rhythm worked in non-vocal music is a matter of
speculation, but some evidence may be extracted from theorists or derived from poetic
sources (see e.g. Phillips 2013).
17
Ar. Ran. 1314, 1349. This kind of comment may be distinguished from obser-
vations (explicit or implicit) about unusual or contrived metrical usage, such as we

In consequence of these and other factors, it is left to modern
interpreters to try to recreate from theoretical statements, documen-
tary indications, and ethnomusicologically supported assumptions
how the phonic effects that can be extracted or imaginatively recon-
structed from musical texts may have contributed to the expression of
a particular passage of ancient poetry and song. In this respect, it is
curious to note that little scholarly effort has been expended on
explicating even such well-known aspects of ancient music as its
rhythms in relation to whole texts or extended passages of poetry,
in the way that literary scholars and philologists standardly offer
‘close readings’ of such texts, or that students of classical music
might provide a bar-by-bar analysis of a sonata or a song. This
comparison draws attention, however, to the repetitive quality of
ancient rhythms and the subtle differences between similar metrical
patterns; and it highlights the difficulty of finding an adequate non-
technical vocabulary to describe the effects of rhythm and melody,
both in general and specifically in the ancient context. Since our task
is not simply to illustrate the way music was attached to the texts but
to ask what difference it made to the words, we may be required not
only to pay close attention to subtle rhythmic and melodic differences
but to find a new repertoire of descriptive terms.
MUSICAL EFFECTS IN HOMER
Homer’s exploitation of the rhythmical features of the hexameter

was recognized in antiquity and highlighted by ancient commenta-
tors.18 However, the melodic features and effects of the sung epic can
be reconstructed only by conjecture. M. L. West’s pioneering hypo-
thetical elaboration of ‘the singing of Homer’ was based on the
following premises:19 (1) the melodic contour of the vocal line fol-
lowed the pitch inflections of Greek words, as recorded by Hellenistic
accent-marks (including some anomalous accents that may be a
find in Hephaestion (Ench. 6) in relation to Sophocles fr. eleg. 1 and Critias fr. 4, and
in Arist. Poet. 1458b5–15; D’Angour (1999) 123–5.
18 19
See n. 7. West (1981) 115–16, 121–2.
memory of sung notes);20 (2) the singer’s voice was accompanied note
for note by a four-stringed phorminx with strings tuned to corres-
ponding pitches;21 and (3) the end of the verse invariably allowed for
a breathing-space—and for the oral composer, a thinking-space—
during which the strings of his lyre could be strummed.22 Extrapo-
lating from what we are told about the structure of Greek modal
systems, West proposed that melody of the song would have utilized a
sequence of four pitches that is found at the core of nearly all the later
systems, to which the strings of the bard’s instrument would have
been tuned.23
The noteworthy subsequent analysis by Stefan Hagel of the inci-
dence of pitch accents in Homer and other early epic poetry suggests
that singers regularly aimed for a rising melodic shape in the early
part of the hexameter verse and a falling melodic cadence at the end.24
Hagel’s tabulations show, in broad terms, that the melodic line
standardly rises at the start, falls at a point around the central part
20
The general accord of melody with word pitch is well attested in the majority of
the surviving musical documents (which preserve other strands of musical tradition as
well as this), and paralleled in cultures with pitch-inflected languages. The fact that the
musical documents all date from post-classical times, and that the earliest papyri with
music from Euripidean tragedy do not show accord with word-pitch, need not cause
problems for this premise (see D’Angour 2006a: 279–80); there will have different
melodic traditions from early times, for some of which (including solo instrumental
and fixed-melody pieces) word pitch would have been irrelevant. See also Franklin
(this volume).
21
In addition to the parallels adduced by West, a good example of this practice
may be found in traditional Ethiopian music using the krar (< kithara), where singers
accompany themselves note for note and intersperse verses with strumming.
22
West (2011) 137 compares Yugoslav oral practice, particularly one recording in
which the singer ‘rests his voice at the end of each verse, even when there is no
syntactical pause’. This is less likely to have been the case with Homeric epic
performance, where the melodization at verse-end was potentially more variable; in
the former case, West (ibid.) notes, that ‘there is almost always a fall on the final
syllable, most commonly of a fifth’.
23
West (1981) 123 gives a transcription in staff notation of the opening lines of the
Iliad as they may have been sung by the bard, with the pitches specified (using
conventional Western musical pitch appellations) as e f a d 0 ; he has subsequently
speculated that (‘to limit the whole compass to a fifth’) the pitches could have been d e
f a (West 2011: 135). For the purposes of our discussion the exact notes are unim-
portant; and it is also clear that some instruments used in Homer’s time would have
had more than four strings.
24
Hagel (1994); graves and acutes are considered equivalent for melodization, but
circumflex accents, where the pitch rise was followed by a longer fall, are not. If both
the latter required distinct melodic treatment from acutes, the technique of melodiza-
tion will have produced effects of considerable refinement.

of the verse, then resumes a rise before a final cadential fall at the end
of the verse. Far less frequently verses end on a high-pitched syllable;
and in such cases, the following line tends to show a falling pitch
around the third-foot caesura, allowing a parabolic arc of melodic
phrasing that rises to the end of one verse to be completed halfway
through the following verse. Since the occurrence of these features is
statistically significant, they show that the way the text of epic was
melodized is not a purely random effect of pitch accord, but one that
will have involved a degree of conscious musical artistry aimed at
shaping the melodic contours of the verse. Hagel’s remarkable study
and findings have yet to be fully appreciated and exploited by
scholars;25 but one immediate implication is that the singer might
manipulate the pitch register at the end of a verse to emphasize, for
instance, a significant word or idea, or might combine a prominent
pitch at verse-end with a falling pitch contour at mid-verse to impart
a melodically calibrated structure to a passage of song.
Accordingly, while one should not suppose that the precise
melodic shape of every verse was the result of deliberate contrivance,
Hagel’s demonstration that the singer may have sought to create
discernible melodic effects demands a closer look at the Homeric
text to see how specific words and phrases might have been enhanced
by their placement in the verse and the resulting melodic phrasing.
Observation of the accentuation found at verse-end in the opening
lines of the Iliad provides a striking impetus to such an investigation:26
μῆνιν ἄειδε θεὰ Πηληϊάδεω Ἀχιλῆος
οὐλομένην, ἣ μυρί᾽ Ἀχαιοῖς ἄλγε᾽ ἔθηκε,
πολλὰς δ᾽ ἰφθίμους ψυχὰς Ἄϊδι προΐαψεν
ἡρώων, αὐτοὺς δὲ ἑλώρια τεῦχε κύνεσσιν
οἰωνοῖσί τε πᾶσι, Διὸς δ᾽ ἐτελείετο βουλή, 5
ἐξ οὗ δὴ τὰ πρῶτα διαστήτην ἐρίσαντε
Ἀτρεΐδης τε ἄναξ ἀνδρῶν καὶ δῖος Ἀχιλλεύς.
Here the first four verses end with on falling melodic contour, so that
when an oxytone first occurs, at the end of the verse in line 5, it does
25
For a concise account of the practical application to Homeric singing, see Danek
and Hagel (1995) with their sung realizations at https://www.oeaw.ac.at/kal/sh/.
26
The importance of verse-end melody may find corroboration in the number of
anomalous Homeric accentuations that place a prominent accent (acute or circum-
flex) on the final syllable; of the nine examples given by West (2011) 13, all but two
have anomalous final accents.
so to salient effect: the phrase Διὸς δ᾽ ἐτελείετο βουλή (‘and the will of
Zeus was fulfilled’) is of clear programmatic significance for the epic.
The three identical falling cadences at the ends of lines 2–4 could be
sung, say, to the descending notes a f e (as might the final two
syllables of Ἀχιλῆος in line 1); the fifth verse would have emerged
with particular emphasis if the singer’s voice and accompanying lyre
notes ended it, in clear distinction to the previous lines, with a
conspicuous rising phrase (e.g. d a) or a series of rising pitches (e.g.
e f a on the last three syllables).
Line 5, however, is not the end of the Iliad’s opening programmatic
‘statement’—a term that carries, of course, musical no less than
literary significance. The final closural verse comes at line 7, where
the two protagonists of the epic are formally introduced, side by side
as it were, with honorific epithets—and again with a climbing final
melodic phrase:
Ἀτρεΐδης τε ἄναξ ἀνδρῶν καὶ δῖος Ἀχιλλεύς.
The high-pitched ending of line 5 thus appears to have laid the
ground for a similar melodic sequence of rising notes at the conclu-
sion of the opening statement at line 7. This may indicate that the
pitch at the end of line 5, shortly to be picked up by that at line 7,
effected a kind of anticipatory closure.27 It might have been sung to
the same notes as final closural phrase in line 7, or may have been
distinguished from the latter by using a different sequence of rising
pitches. Either way, the pitch in these two verses will have risen
prominently above the general level of the song, in contrast to the
falling cadences at the end of the foregoing verses, suggesting that the
singer could have been aiming to create a pattern of thematic markers
through small melodic variations.
The melodic character of the Odyssey’s opening verses presents a
striking contrast to the Iliadic opening. While the Odyssey as a whole
exhibits closely similar frequencies of rising and falling verse-ends,28
its opening lines are markedly different in this respect:
ἄνδρά29 μοι ἔννεπε, Μοῦσα, πολύτροπον, ὃς μάλα πολλά
πλάγχθη, ἐπεὶ Τροίης ἱερὸν πτολίεθρον ἔπερσεν·
27
The notion is familiar to students of modern Western music who use terms such
as ‘half cadence’, ‘final cadence’, ‘imperfect authentic cadence’, etc.; Caplin (1998) 45.
28 29
Hagel (1994) 27. For this accentuation see West (2011) 138.
πολλῶν δ᾽ ἀνθρώπων ἴδεν ἄστεα καὶ νόον ἔγνω,

πολλὰ δ᾽ ὅ γ᾽ ἐν πόντῳ πάθεν ἄλγεα ὃν κατὰ θυμόν,
ἀρνύμενος ἥν τε ψυχὴν καὶ νόστον ἑταίρων. 5
ἀλλ᾽ οὐδ᾽ ὣς ἑτάρους ἐρρύσατο, ἱέμενός περ·
αὐτῶν γὰρ σφετέρῃσιν ἀτασθαλίῃσιν ὄλοντο,
νήπιοι, οἳ κατὰ βοῦς Ὑπερίονος Ἠελίοιο
ἤσθιον· αὐτὰρ ὁ τοῖσιν ἀφείλετο νόστιμον ἦμαρ.
τῶν ἁμόθεν γε, θεά, θύγατερ Διός, εἰπὲ καὶ ἡμῖν. 10
The end of the first verse, with its oxytelic (high-pitched final) syllable
on πολλά, does not effect any kind of closure. Rather, the raised final
syllable is noticeable, and picks up the repeated high notes of the
emphatic opening word ἄνδρά. Both the meaning and melodic sali-
ence of πολλά reinforce the thematic importance of ‘multiplicity’,
hammered home by the polyptoton of πολλῶν and πολλά (3–4).
The climbing melody at the end of the Odyssey’s first verse serves
both musically and syntactically to invite a direct continuation into
the second line. Here the pitch drops at the caesura with Τροίης, to
foreshadow melodically, as it were with a ‘half cadence’, the comple-
tion of the song’s opening melodic arc that comes to rest with a ‘final’
cadence in line 2, πτολίεθρον ἔπερσεν. The oxytelic first verse finds no
echo until the end of line 4, where the final oxytone (θυμόν) recalls the
melodization of the opening verse only to serve as a precursor to the
closural cadence at line 5. Unlike in line 2, however, the musical
phrase does not rest on a half-cadence at the midpoint of line 5, but
remains at a high pitch (ψυχήν), coming to rest only at the end of that
verse with νόστον ἑταίρων, a falling final cadence reminiscent of—but
given the paroxytone ἑταίρων, perhaps blunter than—line 2. The
following five lines continue with standard falling cadences of varied
shapes, only for the melodic phrasing to be reversed at the very end of
the cited passage with a prominent perispomenon, marking the
closural, melodically emphasized, request to the Muse at line 10 to
‘narrate [the story] also to us’ (εἰπὲ καὶ ἡμῖν).
It has long been recognized that the Odyssey begins with a notice-
ably different and more vigorously dactylic rhythm than the Iliad.
The difference is signalled by the trochaic caesura in the first line
and the unavoidable enjambment of the first two lines. But in add-
ition to these distinct rhythmical features, we may now observe that
the composer has arranged his phrases at verse-end to create a
melodic shape that is unmistakably and, one might suppose, inten-
tionally distinct from that of the opening statement of the Iliad.
An interpretation of verse-endings that proposes that a pitch rise at
the end of a verse may be used in some instances as an indication of
marked closure and in others as a means of creating continuity into
the following verse risks being over-explanatory. However, the fre-
quency with which an enjambed oxytelic line is followed by one with
a barytone at the caesura partly mitigates this concern. The possibility
arises that, in melodizing enjambed lines, singers will have carried on
for two or even three lines without an intervening breath or instru-
mental flourish, or at least used a pause of lesser duration between
enjambed lines than at the end of syntactically complete verses and
closural phrases (oxytelic or otherwise) where a longer flourish might
have been demanded. One might further speculate that rising
enjambed lines (such as Od. 1–2) were melodized in a different
manner from those with prominent closural accents such as the
acute in line 4 and the perispomenon in line 10.
The general principle of ‘significant melodization’ that emerges from
these examples clearly merits further examination. For this purpose we
may here proceed with thirteen further verses from the Iliad’s beginning:
τίς τ᾽ ἄρ σφωε θεῶν ἔριδι ξυνέηκε μάχεσθαι;
Λητοῦς καὶ Διὸς υἱός· ὃ γὰρ βασιλῆϊ χολωθείς
νοῦσον ἀνὰ στρατὸν ὄρσε κακήν, ὀλέκοντο δὲ λαοί, 10
οὕνεκα τὸν Χρύσην ἠτίμασεν ἀρητῆρα
Ἀτρεΐδης· ὃ γὰρ ἦλθε θοὰς ἐπὶ νῆας Ἀχαιῶν
λυσόμενός τε θύγατρα φέρων τ᾽ ἀπερείσι᾽ ἄποινα,
στέμματ᾽ ἔχων ἐν χερσὶν ἑκηβόλου Ἀπόλλωνος
χρυσέῳ ἀνὰ σκήπτρῳ, καὶ λίσσετο πάντας Ἀχαιούς, 15
Ἀτρεΐδα δὲ μάλιστα δύω, κοσμήτορε λαῶν·
Ἀτρεΐδαι τε καὶ ἄλλοι ἐϋκνήμιδες Ἀχαιοί,
ὑμῖν μὲν θεοὶ δοῖεν Ὀλύμπια δώματ᾽ ἔχοντες
ἐκπέρσαι Πριάμοιο πόλιν, εὖ δ᾽ οἴκαδ᾽ ἱκέσθαι·
παῖδα δ᾽ ἐμοὶ λύσαιτε φίλην, τὰ δ᾽ ἄποινα δέχεσθαι, 20
ἁζόμενοι Διὸς υἱὸν ἑκηβόλου Ἀπόλλωνα.
At no point in this passage do we find a fully closural melodic leap
such as we did earlier at line 7; rather, the whole passage is notable for
enjambment, and the syntactical period does not come to rest finally
until line 21. Melodic prominences on the final syllable of the last
word of the verse, as indicated by an acute, grave, or (less frequently)
circumflex, occur in six lines (as underlined above), all of which are
followed by verses in which the syllable at the third-foot caesura is
non-accented (also underlined).

Once again, the high pitches of different verse-ends may be seen to
serve more than one function. Thus, the oxytelic line 9 both suggests
an expressive ‘rising’ accentuation of χολωθείς (‘roused to anger’) and
invites continuity through to the melodic fall on ὄρσε in line 10. The
oxytone at verse-end of line 10 on λαοί, which comes at a point of
non-closural pause, keeps the melodic tension of verse-endings at a
high pitch. At the end of 12 comes the first, and melodically prom-
inent, mention of ‘Achaeans’, whose oxytelic form (Ἀχαιῶν) is picked
up by the oxytelic polyptoton (Ἀχαιούς, 15, Ἀχαιοί, 17) a few lines
further on. The words form part of three successive oxytelic lines
(15–17), where the melodic prominence of the terms that spell out, in
different ways, ‘the army of the Achaeans’ (λαὸς Ἀχαιῶν) makes this
referent more of a thematic focus in the passage than would the verse-
end anaphora alone.
What provisional conclusions might we draw from this prelim-
inary investigation of the possible effects of the melodization of
Homeric song? Folk musical melodies universally tend to have an
outwardly repetitive character and limited melodic range, and these
tendencies are likely to have been exaggerated in works of unusual
length such as Homeric (and other) epic.30 The fact that Greek epic
song is rhythmically characterized by the repetition of near-
identical hexameter lines in succession does not, however, prevent
the oral poet from having created a vast range of subtle effects by
varying both intentionally and adventitiously the rhythmical
resources available to him. Pending further statistical and practical
examination, one might venture to claim that the epic singer
deployed melodic phrasing with no less variability than he did
rhythm, so as to make a discernible difference to his performance
in at least three areas: to signal moments of special significance in
his narrative, to reinforce or differentiate the syntactic connection
between successive verses, and to impart a thematic substructure of
melodic echoes to individual passages and to the overall pattern of
his song.
30
Monotonous, repetitive melodies and limited range are evident in Parry and
Lord’s recordings of Yugoslav epic song (excerpts are on the CD accompanying Lord
(2000)); but how these songs sound to modern Western ears should not be confused
with the way they are heard by those familiar with the living tradition.
THE CHORUS OF EURIPIDES
The centuries that followed Homer saw the introduction to Greece of

new musical techniques and traditions. While folk music using con-
ventional rhythms and employing melodic formulas that conformed
to spoken Greek pitch contours must have continued throughout the
period, more exotic, refined, and professional styles of music entered
the repertoire. As lyric song, both choral and monodic, grew in range
and popularity, the singing of Homer yielded to rhapsodic perform-
ances of the epic. Professional citharodes performed passages of
Homer to their own lyre accompaniment, but the absence of the
instrument in accounts and images of rhapsodes indicates that the
original melodic realization of epic was being set aside in favour of
declamation. It may be that, among the plethora of innovative
melodic and rhythmic styles, the subtler effects of Homeric melodiza-
tion were no longer heard or appreciated by listeners for what they
were, and audiences’ engagement with the dramatic narrative eclipsed
the appeal of what may increasingly have seemed a monotonous and
repetitive form of melodic expression.31
The most revolutionary change in musical style, the so-called New
Music, was felt to have taken place between the middle and the end of
the fifth century.32 One key aspect of this revolution was the per-
ceived violation of traditional styles of instrumental and vocal expres-
sion, an eventuality that has been linked with the need for progressive
melodists such as Euripides to find a way, for the first time in Greek
musical history, to notate exactly how a song should be sung.33 It is
hard not to see the influence of these developments on the earliest
surviving substantial fragment of ancient Greek musical notation,
Vienna papyrus G2315, which preserves a few words from a chorus
of Euripides’ Orestes of 408 BC accompanied by both melodic and
rhythmical markings. The melodic line, which is preserved with
both vocal and some interspersed instrumental notation, is notably
31
The kitharodes who were associated with the singing of Homeric passages were
evidently tarred with the same brush: Power (2010) 197, 237. See now also Franklin
(2016) on Stesander.
32
In D’Angour (2006a) I argue that the ‘revolution’ may not have been as sudden
or as radical as is suggested by some of the sources.
33
‘D’Angour (2006a) 282.

adventurous, and is generally thought to have been composed by
Euripides himself.34
In the antistrophe of the first stasimon of Orestes, the chorus
conveys its distress to Orestes by singing in the agitated rhythm of
dochmiacs. I present the relevant passage below in its standard
colometry (which differs somewhat from that of the musical papyrus),
using square brackets to mark the lacunae and showing the melodic
prolongation over two syllables of ἐν (345) and ὡς written out as on
the papyrus (ἐ-εν, ὡ-ως). My translation aims, with some inevitable
awkwardness, to match the position of significant phrases. I use bold
print in both English and Greek to indicate moments where a high-
pitched melody is notated, and a double underline where the notation
indicates a falling melodic cadence. Here, then, the Chorus voices
alarm over the avenging spirits of the murdered Clytemnestra, one of
whom, it sings, is staining the royal house with:
your mother’s blood—which makes you leap in frenzy!
Great good fortune is not lasting for mortals;
I lament, I lament.
Up like the sail 342
of a swift ship, some god shaking
overwhelms it in fearful troubles, as of the ocean’s
rough and deadly waves, in its billows.
ματέρος [αἷμα σᾶς, ὅ σ᾽ ἀναβ]ακχεύει;
ὁ μέγα[ς ὄλβος οὐ μόνιμο]ς ἐν βροτοῖς:
[κατολοφύρομαι κατολο]φύρομαι.
ἀνὰ [δὲ λαῖφος ὥς 342
τι]ς ἀκάτου θοᾶς τινά[ξας δαίμων
κατέκλυσεν δ[εινῶν πόνω]ν ὡ-ως πόντ[ου
λάβροις ὀλεθρ]ίοισιν ἐ-εν κύμα[σιν.
It is visually evident from the above how significant melodic
moments of high pitch or falling cadence are distributed in the
antistrophe. On closer inspection one finds that the final three syl-
lables of line 339 (above in bold) are all set to a single repeated high-
pitched note (e); the italicized final segments in lines 341 and 342 are
set to a lower-pitched falling cadence (b – a♯ – a); the first three
syllables of ὡ-ως πόντου are set to a high-pitched turn (e d e); while
34
This is cautiously assumed by Pöhlmann and West (DAGM 3, p. 16); some
scholars, including Rocconi (2003) 71, are less confident about the Euripidean
provenance of the melody.
the first three syllables of ἐ-εν κύμασιν in line 342 describe a swooping
arc from low to high and down again (a♯ – e – b♭).35
What is the rationale for this melodization? The same melodic
notation evidently accompanied both strophe and antistrophe, and
the melody shows no consistent conformity to Greek word accents.36
This is not simply a function of strophic composition, which can
accommodate a melodic line subject to the same harmonic structure
(similar to the structure of notes used for Homeric singing) to
produce a repeated melody with minor variations across verses and
stanzas.37 Here, however, the melodic line was clearly through-
composed, and on a fundamentally different principle.
The rationale emerges from the way the earlier verses in the
strophe (322–8) are melodized.38 Composed prior to the antistrophe,
it will obviously have generated the original melodization for both
passages:
ταναὸν αἰθέρ᾽ ἀμπάλλεσθ᾽, αἵματος
τινύμεναι δίκαν, τινύμεναι φόνον,
καθικετεύομαι καθικετεύομαι,
τὸν Ἀγαμέμνονος 325
γόνον ἐάσατ᾽ ἐκλαθέσθαι λύσσας
μανιάδος φοιταλέου. φεῦ μόχθων,
οἵων, ὦ τάλας, ὀρε-εχθεὶς ἔρρεις
you who tread the spacious air, her life-blood’s
penalty repaying, repaying murder,
I beseech, I beseech: 325
35
The notation is best interpreted as in the enharmonic, not the chromatic, genus
(see DAGM 3, p. 16), even though Ps.-Psellus On Tragedy 5.39 states that Euripides
diverged from previous tragedians in his use of chromatic (see D’Angour (2017), 436).
36
DAGM 3, p. 16. The papyrus shows line 341 transposed to just before line 339,
which affects how the melody might be interpreted; I take the transposition to be a
scribal error (due to confusion arising from the melodic notes Π Ρ Σ repeated at the
end of two lines) and have interpreted accordingly. Either way the overall relationship
with pitch accents remains the same.
37
Thus it is not the case that ‘the fragment enables us to answer the much-debated
question whether strophic lyric was subject to melodic as well as metrical responsion’
(DAGM 3, p. 16). This fragment can only answer for itself, and may even be evidence
for notated ‘through-composition’ being a Euripidean innovation; D’Angour (2006a)
280–1. For further discussion of Euripides’ melodic practices see Thomas (this
volume).
38
I depart from the transcription in DAGM (p. 13) in supposing that the scribal
error (see n. 36 above) requires us to return line 341 to its received position,
transferring the melodic notation along with the misplaced text.
let Agamemnon’s
son be allowed to forget the frenzies
of mad affliction. Oh for the toils
which you, poor man, stretching out for are lost.
The melodization here nicely supports the testimonies that point
to the way Euripidean musical practice sought to enhance the dra-
matic impact of words by being imitative or expressive of words
and emotions. Such imitiation seems evident in the falling cadence
(b - a♯ - a) to which the last three syllables of the word for ‘I lament’
(κατολοφύρομαι, 341) are set; and this corresponds to the identically
shaped and affectively similar ‘I beseech’ (καθικετεύομαι, 324) in the
strophe. These words both follow phrase-endings which use the same
falling melodic cadence to accompany the phrases ‘for mortals’
(ἐν βροτοῖς, 340) and its strophic counterpart ‘[repay]ing murder’
([τινύμε]ναι φόνον, 323). While the expressive function of the melody
is not so evident in these phrases, the successive repetitions of the
same cadential phrase, creating an aural reminiscence of the melody
attached to ‘I beseech’ and ‘I lament’, serve to emphasize the dejected,
lamentatory impression of the chorus’s sentiments.39
The fact that a modern ear shares a sense of the dejection indicated
by a falling cadence is striking evidence for considering Greek music a
true ancestor of the Western musical tradition. It cannot be taken for
granted that the shape-symbolism perceived by the modern ear was
the same for the ancients; but while one must be cautious of import-
ing modern reactions, there are grounds in this instance for acknow-
ledging a historical continuity in the symbolism of melodic shape.40
Equally noteworthy is the possibility that the composer intended
the thematically important word ‘(her) life-blood’ (αἵματος, 322) to
receive special emphasis from its melodic expression. The word’s
three syllables respond to the latter three of the antistrophe’s ἀνα-
βακχεύει, ‘(leap) in frenzy’, which are set as a group to the highest
note in the fragment, e;41 but the latter has three long syllables, while
39
In my discussion below of the Seikilos song I return to the question of the
continuity of Western melodic shape from ancient music through to Gregorian chant
and beyond.
40
Langer (1976) 226–32 suggests features of the symbolism of musical shape
(without noting that it may be culturally specific to Western musical experience).
41
If the alternative transcription is assumed as in DAGM (see nn 36 and 38 above),
the main corollary is that the high e falls at the end of τινύμεναι φόνον rather than
ἀμπάλλεσθ᾽, αἵματος. While φόνον might be construed no less dramatically than
αἵματος is long-short-long (this more compact word will more easily
have attracted the single pitch than the responding longer one, which
might have been more likely to invite some pitch variation). More-
over, to express the sense of ‘leap’, this high note appears to represent
a melodic upward leap of a large interval, perhaps of a fifth.42
A figure involving a similar upward rise from a♯ to e, followed by
an immediate fall by a fourth to b♭, is imitatively used to melodize
the effect of Orestes’ ‘stretching out’ (ὀρε-εχθείς, 328) in the strophe.
The text of the antistrophe to which this melodic expression attaches
(ἐ-εν κύμασιν ‘in the waves’, 345) offers no imitative rationale, sup-
porting the presumption that this distended, ‘stretched-out’, melodic
figure was originally designed for the strophe. Two other features of
the melodization of ὀρεχθείς are notable. First, the fact that the final
accented syllable is pitched higher than the first syllable allows for a
degree of pitch conformity, by-passing the ‘stretch’ effect of the
intermediate syllable, to be felt. Secondly, the melodic rise on the
second syllable creates an aural expectation that the singers are about
to address Orestes by name, since in terms of sense Ὀρέστα, which
rises tonally on the second syllable, might easily have taken the place
of ὀρεχθείς. While the melody makes less expressive sense as an
accompaniment for the subsequent ‘in the waves’, it serves again to
create an aural reminiscence (as in the case of the repeated falling
cadence on ἐν βροτοῖς etc. discussed above) of the melodic figure
associated with ‘stretching out’ (ὀρε-εχθείς) in the strophe.
Another aspect of the compositional process might provide an
explanation for why no imitative melodic emphasis appears to be
placed, as we might have expected, on the words for ‘up’ (ἀνά, 342)
and ‘shaking’ (τινάξας, 343) in the antistrophe. In each case, these
syllables correspond to words in the strophe that are affectively
neutral and would not obviously invite emphatic melodic
expression—the first two short (unaccented) syllables of Ἀγαμέμνονος
(325) and the third (accented) syllable of the word ‘forget’ (ἐκλαθέσθαι,
326). However, in addition to melody there were choric and
αἵματος, its position at the end of the repeated phrase with τινύμεναι, and the
implication of upward movement connoted by the preceding ἀμπάλλεσθ’, are both
factors that incline me to hear αἵματος as having been accompanied by the high e.
42
The lacuna at this point does not allow certainty: while the pitch level of the
surviving text before the lacuna dwells at around a fifth below this note, other
reconstructions (such as that by von Jan, for instance) might propose less dramatic
intervals.

rhythmical resources available to Euripides’ chorus to add impact to
words, and in the case of both ἀνά and τινάξας a syllabic emphasis
might have indicated by the latter means. Support for this notion is
suggested by the precious markings preserved on the papyrus repre-
senting ictus, the beat that gave form to the chorus’s singing and
dancing: a diacritical point (στιγμή) is used to mark the upbeat
(ἄρσις), indicating that the the downbeat (θέσις) fell on syllables
where ἄρσις is not marked.43 What these markings show confounds
any expectation that the beat of the dochmiac metre coincided with
the long elements of the basic metrical pattern.
The modern reader tends to stress the long elements in metrical
patterns, thus reading the basic five-position pattern ˘ — — ˘ — as ‘di
dum dum di dum’ with three stresses (in bold print) as in the
mnemonic ‘the wise kangaroos’; but there are in fact two places
only in which an upbeat is specified by the ἄρσις mark, one on position
1 and the other on position 3, leaving the downbeat to fall (and
arguably to create greater emphasis) on positions 2 and 4–5. This
articulation makes the rhythm subject to two downbeats of unequal
durations, ‘di dum dum di-dum’ (as heard in the mnemonic ‘that ol’
man river’).44 It is uncertain how far the use of an ἄρσις sign at this
period represents a purely rhythmical convention indicating ‘upbeat’
(as it does in our next example, the Seikilos song), or whether it
preserves a genuine record of the movement of singers’ bodies or
limbs as, for instance, they raised up and brought down their arms or
feet in the dance. Either way, it is noteworthy that in the case of both
ἀνά and τινάξας, the first downbeat (θέσις) of the dochmiac pattern in
the verse coincides with the pitch-accented syllables of those words.45
This suggests that, as a substitute for marking the pitch inflection with
a melodic rise, the θέσις could have imparted to those syllables an
emphasis, perhaps both aural and visual, marked by a stamp of feet.
This would add an obvious expressive weight to τινάξας and ἀνά. The
fact that the sense of ἀνά as ‘up’ is marked by a ‘down’ beat is not an
objection to this supposition; a dynamically accented beat may be
43 44
Anon. Bell. 1.85. Cf. D’Angour (2006b) 491–2.
45
We do not know the melodic accompaniment to λύσσας (326) or δαίμων (343),
but it may be noted that in both cases the θέσις would have fallen on these disyllabic
words, effectively coinciding with the paroxytone.
used to give a sense of the ‘raised’ syllable (as appears to be the case in
the Seikilos song discussed below).46
There is no need to suppose that the enunciation of the vocal line
was always tied to the insistently regular rhythm indicated by the
ἄρσις and θέσις. One of the keys to the exciting impact of the New
Music may have been the vigorous interaction between the complex
patterns of rhythms arising from the disposition of the sung words
and a dance-beat dictated by simpler alternations of up and down
beats. Such complexity would have been no less attractive to the
sophisticated composers and chorus-trainers than the redirection of
traditional expectations of word-pitch accordance into different form
of expression by means of vocal or bodily ictus as well as expressive
melodization. The varied and complex interaction of beat, melodic
line, and pitch-accent offered a wealth of possibilities for creative
poet-musicians to enhance their words through musical settings.
THE SONG OF SEIKILOS
Over half a millennium after Euripides composed the exciting,

innovative music for his Orestes chorus, the same symbols of vocal
notation were used to record the short song recorded on the ‘Seikilos
stele’. Dated to around 150–200 AD, the song preserves precious
testimony to the way ancient musical notation, both melodic and
rhythmical, was applied, and to a style of melodization that would
have been familiar to musicians at this early period of overlap
between pagan and early Christian practices. The song represents
not only one of the most complete of ancient musical compositions, it
is also the most accessible to the modern ear. The reasons for this are
worth noting. First, in marked contrast to Euripides’ agitated and
complex dochmiacs, the form of the song (technically a series of
iambic dimeters with syncopation and resolution) falls neatly and
explicitly (given the ictus-marks included above the melodic
46
In addition to the examples in the above footnote, the θέσις coincides with a
number of other pitch-prominences on other important words where the melodic
setting conflicts (or may have conflicted) with word-pitch, e.g. δίκαν, φόνον (323),
μόχθων (327), τάλας, ἔρρεις (328), ἀναβακχεύει (339), etc. It may be significant that
fewer such coincidences occur in the antistrophe: the composer would not have
attempted to replicate in detail the expressive effects used in the strophe.

notation) into two regular four-bar phrases, a pattern that has dom-
inated Western musical phrasing for hundreds of years.47 Secondly,
the melody, which is diatonic throughout and centres on a repeated a,
is both lightly repetitive and artfully varied.
Transcribable into modern notation in the key of D major as
shown below (but with the dominant a acting as a kind of ‘tonic’
note), the Song of Seikilos sits at least as easily with modern Western
harmonic and rhythmical notions as with ancient modal and metrical
theory.48 With the possible exception of the last two falling notes of
the song (a coda whose function I consider below), the general
melodic form is familiar, and in practice somewhat reminiscent of
the melodic effect of Gregorian plainsong:49
Ὅ-σον ζῆς φαί - νου μη-δὲν ὅ -λως σὺ λυ-ποῦ πρὸς ὁ- λί-γον ἐσ - τὶ τὸ ζῆν τὸ τὶ-λoς ὅ χρόνος ἀπαι - τεἵ
The words to which the music is set have been regularly dismissed
by scholars as slight and banal, and the song itself is often described
somewhat pejoratively (e.g. as a ‘ditty’).50 Yet the sentiment it
expresses, though hardly original, represents a timeless maxim, dig-
nified by no less a philosophical system than Epicureanism, which
47
See in general Rosen (1998) 258–78: ‘On the whole, it is clear that by the 1820s
the four-bar period has extended its domain over musical composition’ (261).
48
Solomon’s (1986) painstaking analysis would be even harder for the song’s
original composer to understand than it is for a modern reader. The transcription
here uses the standard key-signature of D major (initially without the barlines that are
standardly used to mark off the phrases). However, the tune is recognizably centred
on a; all that prevents the designation of A major with a as the tonic is the G natural
(though this clearly functions as a subtonic).
49
Were it not for the provenance of the stele and its conformity to obscure
epigraphic and notational conventions (some of which were not widely recognized
until the twentieth century), on the basis of the musical style alone it might be
suspected of being an accomplished nineteenth-century forgery. According to the
Alypian tables the notation is nominally Ionian (or Iastian), whereas the melody itself
is clearly in the Phrygian species, i.e. it can be played entirely on the white notes of the
piano octave that span d-d 0 , taking g as the tonic in place of a. Moreover, according to
classical theory, the scale created out of the disjunct tetrachords E-a and b-e would
correspond to a mode whose tonal centre should be expected to be b; were this in fact
the case the opening ‘fifth’ would not be a perfect fifth (as it is clearly intended to be,
based on the tonic a) but an irrational interval slightly greater than a fifth.
50
West (1992b) 301.
bears and attracts repetition in every generation.51 Equally it is clear,
as I will show in detail below, that the melody itself has been com-
posed and presented with close care and attention. The elegiac coup-
let that serves as an epigraph to the song sets out Seikilos’ proud claim
to have set up the stone as ‘long-lasting sign of eternal memory’
(μνήμης ἀθανάτου σῆμα πολυχρόνιον).52 It says much for the com-
poser’s musical skill that he makes his melody conform almost
entirely to word pitches without ever allowing these to restrict in
any discernible way the overall shape and gently alternating patterns
of his melodic phrasing. But there are expressive melodic elements in
this apparently slight composition which have been overlooked by
scholars in their zeal either to patronize it or to subject it to excessive
technical analysis.53
A piece with so consistent an accord with word pitch inevitably
draws attention to the one or two occasions on which it diverges from
that accord, and the most obvious example is the rising fifth with
which the song begins. This has been explained as a ‘conventional
incipit’, but the sole ancient parallel adduced for such a practice is
the opening rising fifth of Mesomedes’ short ‘Hymn to the Muse’
(ἄειδε Μοῦσα) composed in the earlier part of the second century
54
AD. One cannot comfortably posit a convention on so slim a base of
evidence. Another explanation is available if we recall the use made by
Euripides of ictus to represent what would otherwise be heard as a rise
in pitch on an accented syllable. It makes perfect aural sense here for
the first syllables of ὅσον and ἄειδε to bear a dynamic stress in place of
a melodic heightening.55 The same use of dynamic rather than
melodic representation might be observed in the case of the other
51
Cf. Horace’s carpe diem (C. 1.11.8). As I write, the pop group ‘Take That’
perform their song Shine to international audiences, while the refrain in Sean
Lennon’s Sunshine Lyrics is another pertinent example.
52
σῆμα πολυχρόνιον sets up a slight tension with μνήμης ἀθανάτου (‘immortal
memory’); the stone will eventually decay, but the memory will last forever. The
musical author may also be playing on the technical meaning of χρόνος, a rhythmical
beat, suggesting that the song to come contains many such χρόνοι, until ‘χρόνος itself
brings it to a close’ (a metapoetic gloss on the final phrase).
53
After twenty-four pages of technical analysis, Solomon (1986) describes it as ‘an
“attractive melody” not without melodic inspiration’ (479).
54
DAGM 24.
55
If ‘the change from a primary pitch accent to a primary stress accent was . . .
widespread by the middle of the second century BC’ (Horrocks (2010) 111), it will have
been well established by the time of Hadrian (whose freedman Mesomedes was).

apparent violation of pitch accent, in respect of ἐστί in the third line.
The graceful melodic figure on the first long syllable incorporates a
slight rise in pitch, but the ictus on the second syllable makes a more
definite impact on the word’s accentuation—which at this period,
moreover, would have been heard in spoken Greek as a dynamic
stress (a point to which I will return).56
Using the opening a as a melodic centre, Seikilos establishes with a
rising fifth to e the harmonic framework of the melody to follow; out
of its thirty-seven notes, only four fall below the ‘tonic’ a and none
rise above the e. The rising interval of a fifth not only embraces the
song’s central melodic span, it imitatively ‘spans’ the word ὅσον—‘so
long as’—in a manner reminiscent of the imitative use of the interval
by Euripides in the melodically distended setting of ‘stretching out’
(ὀρε-εχθείς).57 Further expressive uses of melodic shape are evident
throughout the song. At the end of the second phrase, λυποῦ ends
with a falling melodic figure that symbolically imitates the sense of
despondency inherent in the word’s meaning. The final note of that
phrase, the ‘subtonic’ note g, is melodically a half-cadence; so the
listener grasps that the statement is not yet over—indeed, that we are
only half-way through it. The third phrase illustrates the notion of life
being for a ‘little’ span by beginning with a series of ‘little’ (i.e.
resolved short) syllables (πρὸς ὀλίγον); while the last word of that
phrase, ζῆν, precisely echoes the melody of the final syllable of the
previous line and similarly ends on the subtonic g, reinforcing the
sense that ‘the end’ is yet to come.
The final verse, which begins with the words τὸ τέλος, ‘the end’, is
set to a series of notes that strike the ear as a simple rearrangement of
those in the penultimate verse; yet they are subtly different, with the
addition of a further infixed note (b). This allows for a stepwise
upward progression of a pattern which maps the melodic shape of
τὸ τέλος onto that of ὁ χρόνος at a higher pitch, melodically asserting
that ‘time’ is the arbiter as well as the grammatical subject of the
sentence. The final word ἀπαιτεῖ is melodized to strike a note of
finality, even of doom. In contrast to the high, optimistic sentiments
56
This suggestion obviates the need to posit a paroxytone accent for ἐστὶ (i.e. ἔστι),
as assumed in West (1992b) 3–1 and suggested in DAGM p. 90.
57
The sound and shape of the rising hoson has drawn comparison to the early
chant Hosanna Filio David: Reese (1941) 115. Curiously, the surviving musical
documents preserve other occasions on which a hosanna may be heard in some
form, including the Euripidean ὅ σ᾽ ἀναβακχεύει.
indicated by the high pitch and rising melody of φαίνου in the first
verse, the ‘demand’ of time with its falling final pitches (to the lowest
note of the whole piece, E) brings the hearer to a sober realization of
the inevitability of ending.
The subtly imitative qualities of the melody are masterly, and never
intrude on the musical integrity of the song. The artful compactness
of the song’s form also merits consideration, and our recognition of
the changed pronunciation of Greek in the Roman era alerts us to a
formal feature that has not been generally remarked on: the last
syllable of each couplet of the four-line song is composed to give
the effect of a rhyme. Pronounced correctly for their time, the words
ζῆν ([zi:n]) and ἀπαιτεῖ ([ape:ti:]) at the end of the penultimate and
final verses would have been heard, no less than φαίνου and λυποῦ in
the opening verses, to create an unmistakable assonance. This feature
significantly highlights the remoteness of this composition, with its
rhyme scheme AABB, from classical poetic practice, where rhyme is
never used in this way. Despite detectable elements of expressive
continuity with earlier music such as we have mentioned, this alien
intrusion on classical norms warns us that the use of the musical
notation alone should not mislead us into supposing that the song
was heard to operate according to traditional rhythmic canons any
more than it conforms to classical harmonic theory.
Under these circumstances, it seems as anachronistic to analyse the
song in terms of ancient musical theory and metrics as it would be to
explain an artist’s choice of colours of an ancient mosaic in terms of
spectrographic frequencies. If the rhythm were, in fact, to be straight-
forwardly analysed as ‘iambic dimeters’, the standard scansion of the
text might appear as follows:
Ὅσον ζῇς, φαίνου, ˘——|—— 2 ia
—
μηδὲν ὅλως σὺ λυποῦ· ˘˘—|˘—— 2 ia
πρὸς ὀλίγον ἐστὶ τὸ ζῆν, ˘ ˘˘ ˘ — | ˘ — — 2 ia
τὸ τέλος ὁ χρόνος ἀπαιτεῖ. ˘ ˘˘ ˘ ˘˘ | ˘ — — 2 ia
Setting aside the indications of beat and syllable duration on the stele,
these verses are recognizably iambic in form. The lyric iambic metron
is subject to transformations such as those created by syncopation (a
missing beat, standardly indicated by V) and resolution (two short
beats for one long). Complete with syncopations and ‘bunched’
resolutions, the rhythmical equivalence of each line invites more
appropriate visual representation as follows:
˘—˄—|˄—˄—
—
˘˘—|˘—˄—
˘ ˘˘ ˘ — | ˘ — ˄ —
˘ ˘˘ ˘ ˘˘ | ˘ — ˄ —
Such an analysis, however, alerts us to practical complications. With-
out the στιγμαί to indicate a regularly spaced beat, the second line
could represent (in ‘scansion mode’) the rhythm dum di di dum di
dum [di] dum; that is, it might most easily be read μηδὲν ὅλως σὺ
λυποῦ with stress accents on the long syllables (as underlined) and a
compensatory shortening of the value of the double-short element
(♩. ♬ ♩. ♪ ♩. ♩.).58 In place of this offbeat rhythm, the duration-signs
and στιγμαί show that the intended rhythm was one of evenly pulsed
intervals, μηδὲν ὅλως σὺ λυποῦ (♩ ♪ ♪ ♩ ♪ ♩ ♩). Similarly, in the
absence of στιγμαί the third and fourth lines might more readily be
stressed on the second and fourth elements (with or without reso-
lution) of each iambic metron, i.e. πρὸς ὀλίγον ἐστὶ τὸ ζῆν, / τὸ τέλος ὁ
χρόνος ἀπαιτεῖ. The apparently correct evenly stressed rhythm (πρὸς
ὀλίγον ἐστὶ τὸ ζῆν, / τὸ τέλος ὁ χρόνος ἀπαιτεῖ) may be restored by
showing the staff transciption with barlines (which imply ictus at the
beginning of a bar) in place of the στιγμαί used on the stone:
Ὅ-σον ζῆς φαί - νου μη-δὲν ὅ -λως σὺ λυ-ποῦ πρὸς ὁ- λί-γον ἐσ - τὶ τὸ ζῆν τὸ τὶ-λoς ὅ xρό-νος ἀ -παι-τεἵ.
The result of reinforcing the ictus on the first syllable of each verse,
however, is that the song is easily heard (particularly the last two
phrases, as its nineteenth-century editor Carl von Jan perceptively
noted) as falling into a trochaic rhythm i.e. — ˘ — ˘, a ‘falling’ rather
than a ‘rising’ rhythm.59 Moreover, the substitution of a choriamb
(— ˘ ˘ —) for an iambic metron in the second line represents a rare
metrical variant in classical verse (technically an ‘anaclastic’ iamb, in
which the position of the first two elements ˘ — are reversed to — ˘).60
Its presence here raises further questions about how securely the song
can be analysed in classical metrical terms. The regular ictus and
prolonged notes (it is better to speak of prolongation than the
58 59
Cf. West (1982) 23–4. Jan (1962 [1895]) Supplement p. 36.
60
A possible (but not universally accepted) example of such anaclasis may be
found in the first line of the ‘Nestor’s Cup’ inscription (CEG 454).
traditional ‘syncopation’, which strictly speaking connotes a ‘gap’ in
the rhythmic flow) suggest a different approach to rhythmicization
from that of classical verse, albeit one which has elements of continu-
ity. Rather than subject the song to heavy-handed metrical analysis,
therefore, we might be better advised to trust the aural impression
that the song produces, which is one of syllabic and phrasal balance
reinforced by assonance, comparable to a familiar nursery rhyme:
Humpty Dumpty sat on a wall
Humpty Dumpty had a great fall;
All the king’s horses and all the king’s men
Couldn’t put Humpty together again.61
Finally, however, one might wonder how the increasing dynamic
stresses of the word accents in regular spoken Greek of the time
(touched on above in the discussion of the melodization of ἐστί in
the third phrase) might have interacted with the metre and phrasing.
This becomes particularly acute in the fourth phrase of the song,
where the dynamic accents of second-century speech would have
fallen on syllables that do not coincide with the ictus implied by the
στιγμαί. Taking dynamic stress solely into account, it would be more
accurate to transcribe the latter part of the song with barlines placed
immediately prior to the words τέλος, with the definite article τό
acting as a kind of upbeat before the barline, i.e.:
τὸ τέ - λος ὁ χρό - νος ἀπ - αι - τ - εῖ
Given the natural placing of stresses on the words in spoken Greek at

the time of the song’s composition, this may well be how the enunci-
ation was perceived by the composer, and inscriptional indications
may support this view: the placing on the stele of the στιγμή over the
second syllable of λυποῦ in the second phrase has been ‘corrected’ by
recent editors so that it falls in the centre of the diseme sign.62 This
61
Such a comparison may in the past have led to the song’s being dismissed as a
musical trifle. It is instructive to sing the Seikilos melody to the words of ‘Humpty
Dumpty’, noting how the word ‘fall’ coincides with a melodic fall, the melody speeds
up with ‘all the king’s horses’, and the coda of the final bar strikes a note of dejection
appropriate to the conclusion. While these melodic coincidences may be the result of
felicitous chance, the effects are noticeable.
62
DAGM p. 88, line 7; similar assumptions may have guided the editors’ placing of
the στιγμαί over ζῆν and ἀπαιτεῖ.

leads to an even and regular rhythmical interpretation of the verse, as
shown in the staff transcription above. On the stele, however, the
στιγμή is placed above the ῦ of λυποῦ. This may suggest that the author
considered the ictus as falling on the second note to which that
diphthong is set, as it would be if one sings the words in ‘scansion
mode’ as earlier suggested (i.e. stressed on the long beats as μηδὲν ὅλως
σὺ λυποῦ):
μη - δὲν ὅ - λω-ς σὺ λυ - π - oῦ
In practice, it is open to listeners to switch their perception of ictus

to either manner of execution.63 But when one attempts to sing the
piece according to such dynamic accentuation while simultaneously
preserving the notated ἄρσις, the resulting cross-rhythms give the
performer a different (and arguably more interesting) sense of
melodic movement from that achieved by stressing the words solely
according to the regular alternation of ictus.
A corollary of this observation is that the composer will have been
less prone than we are to hear the verses as rhythmicized in the
regular sing-song way that may have contributed to modern scholars’
dismissive characterizations of the piece. The rhythmical interaction
between dynamic accent and ictus would not be the only anomalous
feature of the song to modern ears. When the stele was first tran-
scribed, the melodic drop by the interval of a fourth at the end of the
song came as an unwelcome surprise to some modern musical com-
mentators, who would have preferred a more familiar cadential close
with reversion to the initial a. In terms of modern harmonic expect-
ations, it is not unusual for a melody to end on the functionally
dominant note of the scale (e), in this case that of the octave below
the tonic. But here the explanation may be found in terms of the
composer’s attempt to create formal balance: as the opening interval
is a rising fifth a-e, the closing interval matches it by being the falling
interval a-E. As well as having an expressive purpose, then, the
falling notes of the last word form an apt counterpart to the rising
63
This is a phenomenon well known to listeners of classical music; one may for
instance ‘choose’ when listening to a performance of Beethoven’s Moonlight Sonata
(1st Movt) to hear the triplets as stressed on any one of the three triplet quavers, even
if the intended ictus is on the first.
notes of the opening word of the song. With the final phrase creating
a fitting melodic response to the opening rising fifth, the song’s octave
range is brought to completion only with the song’s final note. The
ear easily accommodates the sense of closure afforded by the last note
being the lower extremity of the octave E-e, for which the centrally
placed a has provided the tonic centre throughout.
CONCLUSIONS
The particular qualities of Greek music, as demonstrated in examples

that span nearly one thousand years, may be shown to have consist-
ently affected the way texts were understood, performed, and
received. From Homer through to Seikilos, a remarkable continuity
is demonstrated in the relationship of words to sung texts that
depends not solely on conventional rhythmical patterns and pitch
accents but on the use of melody for expressive purposes. While the
precise mechanisms for the use of melody, rhythm, and musical form
for the purpose of textual enrichment varied widely across the cen-
turies, many of the same characteristics and principles are evident in
the singing of Homer, the choral expression of Euripides, and the
song of Seikilos. They all involve the use of music for the heightening
of the emotional and semantic qualities of the text, the patterning of
thematically important elements, and the structuration of their songs.
All these elements are picked up in what is standardly taken to be
the foundational expressions of modern Western music, the Gregor-
ian chants that are first found in notated form in the ninth century AD.
This brings us back to the question of whether we may validly
attribute to the ancients an understanding and employment of
melodic shape in expressive and thematic terms that are so immedi-
ately graspable by a modern Western ear. The answer seems to be that
the earliest music of Greece exhibits elements of symbolic and affect-
ive, no less than geographical and cultural, continuity with the well-
springs of the Western musical tradition. This demonstrable fact
allows us to return to the songs of ancient composers with renewed
illumination and admiration for the way they used rhythm and
melody to make a difference to their texts.
Words and the Musician

Pindar’s Dactylo-Epitrites
Tom Phillips
Scholars have often acknowledged the extent to which our under-

standing of classical choral performances is constrained by our
limited knowledge of music and dance, yet although the metrical
structures of the poems provide us with a means of at least beginning
to reconstruct the expressiveness of ancient songs, there has been
surprisingly little sustained focus on the intersection between rhythm
and semantics in our extant texts.1 This chapter aims to make good
this omission by addressing two aspects of Pindar’s dactylo-epitrite
epinicians in which metre and meaning combine to striking effect.
I first consider rhythmical sequences with expressive or ‘enactive’
force,2 in which significant interactions occur between rhythmical
form and semantic content at the level of individual words or phrases.
I then look more closely at the interweaving of sound and sense over a
longer span of text, focusing on how stanzaic responsion in the
opening two triads of Pythian 1 informs the poem’s metapoetic
I would like to thank the audiences at the ‘Music and Texts in Ancient Greece’
workshop and conference for their responses to early versions of this chapter. I am
also grateful to Armand D’Angour, Felix Budelmann, and Emily Robotham for their
comments at a later stage. All translations are my own.
1
See however Edwards (2002), and the comments of Dale (1969) 248–58 on metre
and meaning.
2
My use of ‘rhythmical enactment’ to describe this phenomenon is based on the
parallel notion of ‘stylistic enactment’, for which see e.g. Silk (2007). In some of the
cases I discuss, rhythmical enactment will be seen to work alongside other aspects of
style: see e.g. pp. 84–7.
74 Tom Phillips
meditations.3 The aim of these readings is to develop a more granular
understanding both of Pindar’s style and the experience of spectator-
ship. In attempting to meet the challenge of finding a vocabulary
to describe and analyse this aspect of Pindar’s poetics, and to throw
an imaginative bridge across the cultural divide that separates mod-
ern readers from ancient audiences, we may begin from some
instances of ancient scholars’ thinking about poetic rhythm.
ANCIENT TESTIMONIES: OPENINGS

AND PROBLEMS
Perhaps the most celebrated instance of ancient literary critics’ inter-

est in the expressive features of poetic rhythm is Dionysius of
Halicarnassus’ discussion of the narrative of Sisyphus’ punishment
at Od. 11.593–8. He begins by commenting on the expressiveness of
the rhythms and syllabic structure of 594–6 (καὶ μὴν Σίσυφον εἰσεῖδον
κρατέρ’ ἄλγε’ ἔχοντα, / λᾶαν βαστάζοντα πελώριον ἀμφοτέρῃσιν, ‘And
I saw Sisyphus bearing great pains, straining with both arms to move
a huge rock’), arguing that ‘the monosyllabic and disyllabic words
leave many intervals between each other and imitate the time the
action takes (τὸ χρόνιον ἐμιμήσαντο τοῦ ἔργου). The long syllables,
which have a holding, delaying quality, give an impression of the
weight and the difficulty of the deed (αἱ δὲ μακραὶ συλλαβαὶ στηριγμούς
τινας ἔχουσαι καὶ ἐγκαθίσματα τὴν ἀντιτυπίαν καὶ τὸ βαρὺ καὶ τὸ
μόλις)’. Then, describing phrases such as λᾶαν ἄνω ὤθεσκε ποτὶ λόφον
(‘he pushed the rock towards the crest’, 596) he remarks that ‘when
considered in respect of their length, the rhythms portray the strain of
his limbs and the dragging as he rolls his burden along and his
propping up of the boulder’. Sound and sense then combine to
provide a vivid impression of the rock rolling down the hillside in
598 (αὖτις ἔπειτα πέδονδε κυλίνδετο λᾶας ἀναιδής, ‘and then again the
pitiless boulder tumbled down to the plain’): ‘doesn’t the arrangement
3
Relatively little attention has been paid to this phenomenon; an exception is
Mullen (1982) 90–142, but his focus is on the structural role of the epode: see below
n. 18. See also Phillips (2013) 53–5; Sobak (2013). For a catalogue of verbal and
thematic responsions and doxography of previous scholarship on the subject see
Stockert (1969).
Pindar’s Dactylo-Epitrites 75
of the words tumble down with the impetus of the rock, and doesn’t
the speed of the narration even outstrip the rock’s movement’ (οὐχὶ
συγκυλίεται τῷ βάρει τῆς πέτρας ἡ τῶν ὀνομάτων σύνθεσις, μᾶλλον δὲ
ἔφθακε τὴν τοῦ λίθου φορὰν τὸ τῆς ἀπαγγελίας τάχος;).4 Dionysius’
reading testifies to the expressiveness attributed to rhythmical form
and the critical care directed at its elucidation.5
The interrelation of sound and sense is also at issue in a less
known but equally sophisticated analysis of rhythmical affectivity
found in the metrical scholia to O.1. The analysis focuses on the use
of a colon comprising six short syllables (˘ ˘˘ ˘ ˘˘) in the thirteenth
line of the strophe (Σ O.1.metr., i Dr 14):6
The thirteenth [sc. colon] is a proceleusmatic catalectic dimeter. In
which observe a marvellous effect. For this rhythm is appropriate for
swiftness (ὁ τοιόσδε ῥυθμὸς ταχυτῆτι ἁρμόττει), whence its name ‘pro-
celeusmatic’. But look now at how the brilliant Pindar has been found to
reinforce such a rhythm throughout the poem so that it stands out, by
successfully expressing speed in the meaning of his words. (βλέπε τοίνυν
τὸν σοφώτατον Πίνδαρον πῶς ἥλω ἐν οἷς τὸν τοιοῦτον ῥυθμὸν εὔσημον
διὰ τοῦ λόγου ἀποκαθίστησιν, ἐν τῇ ἐννοίᾳ ταχυτῆτα κατωρθωκὼς ἐν οἷς
λέγει): ‘whence the much-spoken . . . ’ (ὅθεν ὁ πολύφατος, 8), because of
the speed of rumour; and ‘mind with sweetest . . . ’ (νόον ὑπὸ γλυκυτάταις,
19) because of the swift movements of the mind; and, ‘when his father
called’ (ὁπότ’ ἐκάλεσε πατήρ, 37), because of the speed of the call; and
‘that into water on the fire’ (ὕδατος ὅτι τε πυρί, 48) because of the ease of
movement of the elements involved, the one rising, the other falling; and
again, ‘among the short-lived . . . ’ (μετὰ τὸ ταχύποτμον, 66); and, ‘me on
the swiftest . . . ’ (ἐμὲ δ’ ἐπὶ ταχυτάτων, 77); and ‘of Pelops, where of
swiftness . . . ’ (Πέλοπος ἵνα ταχυτής, 95), and in addition to these [pas-
sages] ‘a god being a guardian’ (θεὸς ἐπίτροπος ἐών, 106) because of the
swift movement of the divine that exceeds even the dart of the eye.
4
Parker (2007) is sceptical about Dionysius’ reading, pointing out that the metrically
remarkable feature of 598 is not that it is holodactylic, but that ‘all the bicipitia except
one are split by word end’ (p. 298). However, she does not mention the sound pattern of
the line, in which the repetition of τ and δ adds to the effect: although Dionysius makes
no reference to this, its articulation of the line shapes the σύνθεσις to which he responds.
5
Sound effects in language were theorized extensively in the classical period (see
Porter (2010) 308–30), and were crucial to the ‘euphonist’ critics of the Hellenistic
period: for an overview see Janko (2000) 120–89.
6
In Snell–Maehler’s colometry, this element is analysed as iambic as part of a line
consisting of three iambic metra. On the ancient metrical analysis cf. Santé (2008)
58–9. For comments on the metre of O.1 in general cf. Mullen (1982) 178–84; Itsumi
(2009) 142–3.
76 Tom Phillips
According to this commentator, Pindar has cleverly employed verbal
phrases that convey ideas related to speed (ἐν τῇ ἐννοίᾳ ταχυτῆτα
κατωρθωκώς) to fit with the character of the rhythm.7 The scholiast
has pursued the effect across the whole poem, and saw rhythm and
sense as combining to produce a ‘marvellous effect’. The scholium is
also notable for the range of relationships that it envisages between
rhythmical form and semantic content: in some phrases, such as
those of lines 66, 77, and 95, the rhythm reinforces the direct verbal
references to speed, but in others the relationship is more complex. In
νόον ὑπὸ γλυκυτάταις and ὁπότ’ ἐκάλεσε πατήρ, for instance, the
rhythm encourages the listener to impute speed to the concepts and
actions signified, even though this dimension is not obvious at a
verbal level. In these cases, rhythm supplements rather than rein-
forces semantics. By suggesting such complex interactions, the scho-
lium provides a useful indicator of how subtly and variously rhythm
and meaning can be related.8
The primary texts under discussion are especially marked instances
of rhythmical enactment, and we should be cautious about inferring
from these examples that such effects were perceived to be common
by ancient readers.9 Nevertheless, these readings are useful as pointers
to the effects Pindar’s poetry may have generated when performed,
the terms of the analysis prompting us to imagine how such effects may
have been augmented in a choral performance by gesture and instru-
mental setting. Yet the move from analysing rhythmical structures on
the page to thinking about how they may have been bodied out in
performance poses considerable problems.10 The metre of a poem is
not necessarily a sure guide to how it would have been performed,11 as
7
Itsumi (2009) 150–1 makes similar comments, although without reference to the
scholium.
8
Although there is of course no parallel for such a run of short syllables in
the dactylo-epitrite odes.
9
This is the only instance of such comment in the metrical scholia to Pindar.
However, the characterization of the effect θαυμάσιον as does not imply that it is in itself
unusual, but that it is ‘wonderful’ in its sophistication (cf. σοφώτατον). Cf. also Hera-
cleodorus on change in rhythms as changing how an utterance is understood: Phld. De
poem. 1.34 with the remarks of Janko (2000) 223 n. 8. For caution about metrical design
for particular effects see e.g. Dale (1969) 258; West (1982) 39; Parker (2007) 297.
10
In relation to Pindar see Mullen (1982), and more generally the cautionary
comments of Dale (1968) 204.
11
On the debate about whether rhythm and metre would have been identical in
fifth-century poetry see e.g. Naerebout (1997) 202–6. Pöhlmann (1995) argues
strongly for their non–identity. Even if identity of rhythm and metre did predominate,
syllabic quantities are only part of what constitutes rhythm as realized
in performance: syllables can be sung with varying durations, so
altering the tempo of the piece. In his analysis of poetic rhythms,
Aristides Quintilianus states that ‘rhythmical tempo is the quickness
or slowness of durations’ (ἀγωγὴ δέ ἐστι ῥυθμικὴ χρόνων τάχος ἢ
βραδυτής), and his subsequent remarks make clear that such tempi
could vary considerably both in form and effect. This proposition is
further complicated by the role of what the ancients termed ἄρσις and
θέσις, ‘rise’, ‘up-beat’, and ‘placing’, ‘down-beat’ respectively. These
terms derive from the movements of the dancers; θέσις is heavier,
and refers to the ‘placing’ of the dancers’ feet, while ἄρσις refers to
the lifting of the foot.
Aristides’ comments make it clear that these movements played an
important role in articulating rhythm (De mus. 1.13):
Rhythm, then, is a system of durations put together in an ordered form
(ῥυθμὸς τοίνυν ἐστὶ σύστημα ἐκ χρόνων κατά τινα τάξιν συγκειμένων).
We call the modifications of these durations ἄρσις and θέσις, sound and
silence. Notes as such, because of the lack of differentiation in their
movement, leave the interweaving of the melody obscure and confuse
the mind: it is the elements of rhythm that make clear the force of
the melody (τὰ τοῦ ῥυθμοῦ μέρη τὴν δύναμιν τῆς μελῳδίας ἐναργῆ
καθίστησι), moving the mind part by part, but in an ordered
way. ἄρσις is the upwards movement of a part of the body, θέσις the
downwards movement of the same part.
The statement that ‘the elements of rhythm . . . make clear the force
of the melody’ testifies to rhythm’s importance: by altering τὰ τοῦ
ῥυθμοῦ μέρη, that is, by lengthening the duration of syllables, a
performer or performers could alter the ‘force’ of a given metrical
sequence. The use of δύναμις here clearly refers to the ‘ethical’ effects
of different rhythms, a topos in ancient discussions of musical cul-
ture,12 and a subject which Aristides discusses in the second book of
his De musica. Here he presents a detailed explication of the analysis
mentioned above, in which long and short syllables, and also different
kinds of rhythmical arrangements, produce different effects (2.15):
however, it is possible that poets and performers could have introduced a certain
amount of variation for aesthetic effect.
12
See e.g. Ar. Poet. 1447a26–7 with Scott (2005) 157–8.
78 Tom Phillips
τῶν δ’ ἐν ἴσῳ λόγῳ οἱ μὲν διὰ βραχειῶν γινόμενοι μόνων τάχιστοι καὶ
θερμότεροι, <οἱ δὲ διὰ μακρῶν μόνων βραδύτεροι> καὶ κατεσταλμένοι, οἱ
δ’ ἀναμὶξ ἐπίκοινοι. εἰ δὲ διὰ μηκίστων χρόνων συμβαίη γίνεσθαι τοὺς
πόδας, πλείων ἡ κατάστασις ἐμφαίνοιτ’ ἂν τῆς διανοίας. διὰ τοῦτο τοὺς
μὲν βραχεῖς ἐν ταῖς πυρρίχαις χρησίμους ὁρῶμεν, τοὺς δ’ ἀναμὶξ <ἐν> ταῖς
μέσαις ὀρχήσεσι, τοὺς δὲ μηκίστους ἐν τοῖς ἱεροῖς ὕμνοις.
Of the rhythms in equal ratio, those composed from short syllables only
are the swiftest and liveliest, while those which are composed of long
syllables and slower and calmer, and those that are a mixture of both
share in both qualities. If it happens that the feet are constituted out of
very long durations, the calming effect on the mind is greater. On this
account, we observe that short durations are useful in war-dances, mixed
durations in dances of the middle type, and the longest in the holy hymns.
Later in the discussion, Aristides elaborates on the ethical effects of
rhythms by comparing them to different ‘styles of walking’; those who
walk with equal steps of a good length approximate to spondaic
rhythms, which are ‘stable and manly in character’, whereas short
and irregular steps, like the corresponding rhythms, are indicative of
irrationality and dissipation. For Aristides, ‘rhythms whose tempo is
swifter are livelier and active, those in which it is slow are relaxed and
peaceful’ (τῶν ῥυθμῶν οἱ μὲν ταχυτέρας ποιούμενοι τὰς ἀγωγὰς θερμοί
τέ εἰσι καὶ δραστήριοι, οἱ δὲ βραδείας καὶ ἀναβεβλημένας ἀνειμένοι τε
καὶ ἡσυχαστικοί). He regards the dactylic metra as ‘more noble that
all the others on account of always having a long syllable in the
leading position’ (σεμνότερον γὰρ ἁπάντων διὰ τὸ τὴν μακρὰν ἀεί
ποτε καθηγουμένην ἔχειν).13 These comments parallel those of Dio-
nysius of Halicarnassus, for whom the spondee ‘has great dignity and
much solemnity’ (ἀξίωμα δὲ ἔχει μέγα καὶ σεμνότητα πολλήν), while
the dactyl ‘is very elevated and most apt for beauty of expression’
(πάνυ δ’ ἐστὶ σεμνὸς καὶ εἰς τὸ κάλλος τῆς ἑρμηνείας ἀξιολογώτατος).
Dionysius also comments that the epic hexameter owes much of its
formal beauty to its dactylic component (τό γε ἡρωικὸν μέτρον ἀπὸ
τούτου κοσμεῖται ὡς ἐπὶ τὸ πολύ). Despite the generalizing nature of
these comments and their ethical slant, they provide a useful template
for thinking about the affectivity of dactylo-epitrite.14 Although we
13
Aristid. Quint. De mus. 1.24.
14
For general comments on dactylo-epitrite see e.g. Dale (1969) 53–60; West
(1982) 69–76. For convenience, I use ‘epitritic’ to denote the combinations denoted
by E and e in Maas’ scheme. For a definition of ‘epitrite’ see West (1982) 70. The use of
should be wary of assuming that these later critics heard when
reading the same effects that early audiences heard when attending
performances, the rhythms’ affective qualities are unlikely to have
altered so much as to make such reflections entirely irrelevant to
analysis of Pindar’s epinicians as songs in performance.
Yet these passages also highlight the provisionality of our evidence.
In Aristides’ scheme, Pindar’s poetry would be classified as ἀναμίξ,
containing as it does a mixture of cola in which long and short
syllables predominate. The metre’s combination and juxtaposition
of such cola is also suggestive, as I shall argue below.15 Unfortunately,
however, we do not know what kind of ἄρσις and θέσις patterns
Pindar and his choruses employed. We are also in the dark about
the type of ‘durations’ employed by fifth-century performers, and no
simple correlation can be drawn between long syllables and stress
patterns.16 Moreover, attempts to analyse rhythm in Pindar’s epini-
cians are complicated further by the poems’ triadic form and the use
of dance in choral performance. In addition to the role played by
rhythmical sequences in articulating listeners’ understanding of long,
syntactically involved sentences and the development of complex
ideas,17 the final section of my argument will suggest some ways in
which rhythmical responsion between and within triadic structures
can be deployed to aesthetic effect.18 However, even a brief glance at
some of the names of dance figures that have come down to us reveals
this terminology should not lead us to reify these metrical building-blocks as com-
positional tools, or as modalities of reception for Pindar’s fifth-century audiences.
Both Pindar and his audiences are likely to have had a much more intuitive grasp of
rhythmical constructs than would be implied by too systematic a retrojection of
‘dactylo-epitrite’ onto fifth-century compositional and interpretative mentalities.
The value of the term lies in its heuristic benefits rather than in its historical accuracy.
15
Cf. Dale (1969) 251.
16
The musical notation on the papyrus of Euripides’ Orestes shows that long
syllables and θέσις do not always coincide; cf. Pöhlmann (1995), and D’Angour
(this volume) pp. 63–4.
17
A feature of rhythm that has received considerable attention in modern cogni-
tive studies: see e.g Thaut (2005) 5 on rhythm as articulating patterns of meaning:
‘discernible temporal distribution and organization of events in groupings imposed by
a rhythmic structure [are] a way of imposing order and meaning onto the perceptual
process’.
18
The most thoroughgoing attempt to analyse the structural aspect of Pindaric
triads is Mullen (1982). Central to his analysis is the notion of ‘epodic arrest’, based on
the premiss that the chorus were stationary during the singing of the epode, whereas
they moved to the left and right during the strophe and antistrophe. However, the
sources that claim that epodes were sung when the singers were standing still
80 Tom Phillips
the problems posed by our lack of evidence for Pindaric choreog-
raphy. Terms applied to the σχήματα of ancient dances, such as
‘whirl’, ‘Pan’s leg’, and ‘sword-dance’,19 hint at the richly mimetic
dimension of dance, and at the complex performance realities that lie
behind Plato’s analysis of of χορεία in the Laws, in which he argues
that ‘no-one who is using his voice, whether in songs or speeches,
can remain very calm in his body’. This results in the emergence
of ‘mimesis of what is said with gestures’ (μίμησις τῶν λεγομένων
σχήμασι), a practice that ‘constitutes the whole art of dancing’ (τὴν
ὀρχηστικὴν ἐξηργάσατο τέχνην σύμπασαν).20
Plato’s claim for the mimeticity of all dance, and retrojections from
later evidence to the practices of the fifth century, have rightly been
questioned,21 but the potential complexities of the relationships
between words, rhythms, and gestures in performance should be
borne in mind throughout what follows. The most important point
to emerge from the preceding remarks, however, is that metrical
(collected at Färber (1936) ii 14–19, and discussed by Mullen (1982) 228–30) all date
to after the first century AD, and have been considered an invention of that period; see
e.g. Crusius (1888). In these late sources, dance has an astrological dimension,
reflecting the movement of the cosmos (see e.g. iii 306 Dr), but there is no evidence
for this idea in the classical period: on the connection between chorality and the
heavens in early poetry see Ferrari (2008) 2–6. Mullen’s attempts to infer choreog-
raphy from choral language have been tellingly criticized (see Burnett (1984) 156–7),
and some of his inferences, such as that the absence of movement from the epode
would have allowed for the audience to be ‘engaged all the more deeply with the sound
and meaning of the song’ (Mullen (1982) 92) do not stand up to scrutiny. Moreover,
Mullen does not mention a significant piece of evidence against the presence of the
notion of epodic arrest in Hellenistic scholarship. In his discussion of triadic con-
struction at Comp. 19, Dionysius of Halicarnassus makes the point that composers
could alter melody and rhythm in epodes but not from strophe to antistrophe (περὶ δὲ
τὰς καλουμένας ἐπῳδοὺς ἀμφότερα κινεῖν ταῦτα ἔξεστι), but does not mention any
difference in choreography between epode and the rest of the triad. Given that the
differentiation of the epode is at issue, it seems unlikely that Dionysius would have
omitted to mention the change to standing still had it been known to him. This
conclusion further supports the argument that the standing epode was a construct of
later scholarship. The recent application of epodic arrest to Pindar by Sobak (2013)
does not take account of the criticisms levelled at Mullen’s thesis.
19
Recorded at Eust. 1166.10 on Il. 18.590, Hesych. s.v., Ath. 14.629f respectively.
For analysis of σχήματα and a list of further such terms, see Lawler (1954) 151–4.
20
Pl. Leg. 816a. Cf. Peponi (2009) 59–60; Prauscello (2013b).
21
On mimeticity in general see Naerebout (1997) 109, and in peripatetic dance
theory see Scott (2005) 158, who argues that Aristotle did not conceive all dances to be
necessarily mimetic of ‘character’. On the retrojection of later evidence see e.g. Lawler
(1954); Naerebout (1997) 269–73; for an overview of the aesthetics of dance see
Peponi (2015).
structures are best conceived as templates on which performers could
have drawn in order to create musical, rhythmical, and gestural
effects, rather than as self-sufficient phenomena.22 My aim in the
following analyses is therefore to note passages in which significant
intersections between rhythmical and semantic structures may have
been at work, and comment on the possibilities these create for
performance and interpretation. These readings will suggest that
while Dionysius and Aristides provide useful guidelines, greater com-
plexities are likely to have been at work in the relation between
rhythm and semantics than can be captured by the terminology
employed by the ancient critics.
MICRORHYTHMICAL EFFECTS
The remarks of the critics quoted above suggest that epitritic cola
might have sometimes been employed to convey and complement a
sense of weightiness, slowness, or staticity, and there are numerous
examples of such affectivity to be found in Pindar, especially in
conjunction with period or stanza end. An example of a climactic
epitrite with a force which ancient listeners (and readers) are
likely to have recognized as possessing σεμνότης is O.8.44, πεμφθὲν
βαρυγδούπου Διός (— — ˘ — — — ˘ —).23 Here the analectic epitrite
with long in anceps position complements the weightiness and grand-
eur of Zeus’ thunder. At P.1.75, Pindar describes Hiero in a vivid
metaphor as ‘dragging Greece out of heavy enslavement’ (Ἑλλάδ’
ἐξέλκων βαρείας δουλίας). Here the slow epitritic rhythm (— ˘ — —
—
˘ — — — ˘ —) complements both the notion of weightiness and
connotes the effort involved in the ‘dragging’ metaphor. Equally,
dactylic cola often occur in association with forms of movement,
comparable to the scholiast’s analysis of O1s13 described above. At
O.3.3, in the phrase ὕμνον ὀρθώσαις, ἀκαμαντοπόδων (‘raising a hymn
for [horses] with untiring feet’, — ˘ — — — ˘˘ — ˘˘ —), the three long
22
The practical role played by the poet in rehearsing his chorus and overseeing the
implementation of particular dance moves and other performance features is unclear;
see e.g. the comments of Mullen (1982) 12–21; and Wilson (2000) 81–6 on choral
training at Athens.
23
The line is spoken by Apollo, and refers to a ‘vision’ (φάσμα) ‘sent from loud-
thundering Zeus’.
82 Tom Phillips
syllables of ὀρθώσαις offset the dactylic movement of ἀκαμαντοπόδων,
the dactylic movement of which suggests the speed of the galloping
horses. Similar is O.3.26, where a dactylic rhythm coincides with
the description of Leto’s ‘horse-driving daughter’ (Ἰστρίαν νιν· ἔνθα
Λατοῦς ἱπποσόα θυγάτηρ: — ˘ — ˘ — ˘ — — — ˘˘ — ˘˘ —).
The phrasing of ὕμνον ὀρθώσαις, ἀκαμαντοπόδων also exemplifies
another common feature of Pindar’s rhythmical poetics, namely the
contrastive juxtaposition of dactylic and epitritic movements. The
scholium which glosses ὀρθώσαις explains the term as ‘having raised
up, elevated, and exalted the ode’ (ἣ ὀρθώσας καὶ ὑψώσας καὶ αὐξήσας
τὸν ὕμνον, Σ O.3.5c = i 107 Dr),24 and it is easy to see this rhetoric
of exaltation as reinforced by the ‘great dignity and . . . solemnity’
Dionysius sees at work in rhythms based on long syllables. The
affectivity of this syllabic structure is heightened by contrast with
the lighter syllables that follow. I now turn to three longer passages in
which similar juxtapositions occur, in order to illustrate the range of
effects that can be created by these techniques. The first the descrip-
tion in Olympian 3 of Heracles’ visit to the Hyperboreans and his first
glimpse of the olive trees that he will transplant to Olympia (31–2):
—
˘˘—˘˘———˘—
τὰν μεθέπων ἴδε καὶ κείναν χθόνα D–e
— —
˘˘—˘˘—
πνοιαῖς ὄπιθεν Βορέα –D
— —
˘˘—˘˘———˘—
ψυχροῦ· τόθι δένδρεα θάμβαινε σταθείς. –D–e
Pursuing her [sc. the doe] he also saw that land beyond the blasts of cold
Boreas; there he stood and marvelled at the trees.
There are several rhythmically suggestive phrases in this passage.
The light movement of ὄπιθεν Βορέα fits the wind’s motility, while
the spondaic words πνοιαῖς and ψυχροῦ balance each other rhythmic-
ally and tonally, with perispomenon on the second syllable. Singers
might have exploited this feature to imitate the sound of the wind,
24
Cf. Porter (2016) 353, who discusses this passage as one of several in which later
critics reuse Pindar’s own lexis of elevation in order to figure him as a ‘sublime’
author. Catenacci (2013) 416–17 notes that the ‘setting up’ suggests ‘una statua o un
monumento ben visibile’, comparing e.g. I.1.46. This interpretation follows that of
Σ O.3.5b = i 107 Dr (τὸ δὲ ὀρθώσας εἶπε μετενεγκὼν ἀπὸ τιθεμένων ἀνδριάντων ἢ
ἀγαλμάτων). On this reading, the connotations of staticity and weightiness in the
rhythm would complement the verb’s monumental connotations.
with the words’ long syllables allowing the effect to be prolonged.
Differently, the three long syllables of θάμβαινε stand in contrast to
the lighter τόθι δένδρεα, the heavier, drawn out rhythm intimating
both the gravity and temporal duration of Heracles’ ‘wonder’. By
creating an analogy between Heracles’ extended perception and the
listeners’ experience of the phrase that describes it, rhythm formalizes
a moment of contact between two worlds.25
The simile with which Olympian 6 opens, in which the poem is
compared to a building with ‘golden pillars’, provides another
instance of heavier syllables accentuating a moment of significant
perception. The poem’s second line contrasts lighter and heavier
movements: κίονας ὡς ὅτε θαητὸν μέγαρον (— ˘˘ — ˘˘ — — — ˘˘ —).
As with θάμβαινε at O.3.32, the place of the three long syllables of
θαητόν in the time of performance, juxtaposed with the surrounding
dactylic and anapaestic rhythms, accentuates the ‘marvellousness’ of
the object described by making the listener linger over it.26 Also
comparable is N.1.39–42:27
—
˘˘—˘˘—
ἀλλὰ θεῶν βασιλέα D
— — — — — — —
˘ ˘˘ ˘˘
σπερχθεῖσα θυμῷ πέμπε δράκοντας ἄφαρ. –e–D
— — — — —
˘ ˘
τοὶ μὲν οἰχθεισᾶν πυλᾶν E
— — — —
˘˘ ˘˘ ˘˘
ἐς θαλάμου μυχὸν εὐρὺν ἔβαν . . . D d2
But the queen of the gods, angered in her heart, immediately sent snakes. They
went, the doors being open, into the wide recess of the bed-chamber . . .
25
Rhythm here reinforces the expressive force of the syntactical structure. Cf. the
comments of Catenacci (2013) 429 on line 32: ‘il passagio asindetico, la brevità della
frase e la costruzione participiale esprimono lo stupore e l’improvviso arresto della
corsa di Eracle ammirato alla vista degli alberi’.
26
Cf. the responding line in the fourth strophe (ἔνθα οἱ ὤπασε θησαυρὸν δίδυμον,
‘where he gave him a twofold treasury’, O.6.65) where the long syllables of θησαυρόν
are emphasized by contrast, producing an effect similar to that of ὀρθώσαις (cf. n. 25).
27
Cf. also O.6.22–5: ὦ Φίντις, ἀλλὰ ζεῦξον ἤδη μοι σθένος ἡμιόνων, / ᾇ τάχος, ὄφρα
κελεύθῳ τ’ ἐν καθαρᾷ / βάσομεν ὄκχον, ἵκωμαί τε πρὸς ἀνδρῶν / βάσομεν ὄκχον, ἵκωμαί
τε πρὸς ἀνδρῶν / καὶ γένος (‘O Phintis, yoke for me the mules’ strength as swiftly as
may be, so that we may drive the chariot on the clear path, and that I might arrive even
at these men’s lineage’: – E – D/D – d1/D – e –). Here the dactylic rhythms of σθένος
ἡμιόνων and βάσομεν ὄκχον, ἵκωμαί match the ‘swift mules’ and the poetic journey,
while the cretics of τε πρὸς ἀνδρῶν / καὶ γένος create a heavier crescendo which
coheres with the sense of the imagined journey coming to an end.
84 Tom Phillips
The dactyls of πέμπε δράκοντας ἄφαρ and ἐς θαλάμου μυχὸν εὐρὺν
ἔβαν coincide with the swiftness of the movement described, and
frame the epitrite of τοὶ μὲν οἰχθεισᾶν πυλᾶν. The intersection of
rhythm and sense in 41 creates an effect which might be compared
with Dionysius’ description of the ‘holding, delaying quality’ of long
syllables, this quality here reflecting the staticity of the doors in
contrast with the movement by which they are surrounded.28
These passages all involve a reasonably direct connection between the
rhythms’ associative forces and the primary qualities of the referents, yet
there are also numerous passages in which the putative connections
between rhythm and sense are less straightforward, and which caution
against too simple a conception of how Pindar and his choruses may have
constructed and exploited this relationship. Several connections of this
sort are at work in Olympian 12, composed for the long distance runner
Ergoteles of Himera.29 As well as creating a series of complex interactions
between form and meaning, aspects of the poem’s rhythmical structure
also form part of Pindar’s appropriation of Homeric language.
The poem begins with a prayer to Zeus and σώτειρα Τύχα, ‘by
whom swift ships are guided on the sea, and rapid battles on the land,
and assemblies that give counsel’ (3–5):
—
˘———˘———˘—
τὶν γὰρ ἐν πόντῳ κυβερνῶνται θοαί E–e
— — — — — — — —
˘ ˘ ˘˘
νᾶες, ἐν χέρσῳ τε λαιψηροὶ πόλεμοι E – d1
— — — —
˘ ˘˘
κἀγοραὶ βουλαφόροι. e – (D)
Both lines 3–4 describe entities that are ‘swift’, and yet do so in
rhythmical phrases in which long syllables predominate. This general
effect is given additional point by appropriation of Homeric vocabu-
lary and metrical form. First, the phrase θοαί / νᾶες. The adjective is
common in Homer, where it is often used of ships. It occurs five times
in the nominative plural, but always with the noun preceding, as at Il.
11.666 (ἦ μένει εἰς ὅ κε δὴ νῆες θοαὶ ἄγχι θαλάσσης) and Il. 16.168
(Πεντήκοντ’ ἦσαν νῆες θοαί, ᾗσιν Ἀχιλλεύς).30 In Homer, therefore,
28
See p. 74.
29
For an overview of the poem’s wider socio-political context see Nicholson
(2016) 237–61 with further references.
30
Adjective and noun are separated at Il. 2.619: νῆες ἕποντο θοαί, πολέες δ’ ἔμβαινον
Ἐπειοί.
the phrase scans — — ˘ ˘, with the diphthong shortened by correption.
By inverting the Homeric word order, Pindar produces a rhythm
that is likely have sounded unfamiliar to audiences accustomed to
the Homeric phrase. Moreover, on two of the occasions on which
the phrase νῆες θοαί occurs in the Iliad, it is preceded by a verb
in the passive with the same metrical shape as Pindar’s κυβερνῶνται
(˘ — — —): ἠὲ φυλάσσονται νῆες θοαὶ ὡς τὸ πάρος περ (Il. 10.309).31
The difference in context is notable: whereas Pindar’s line occurs as
part of a prayer that affirms σώτειρα Τύχα, Homer’s line is a question
about the state of the Greek camp and ships, spoken first by Hector
and then by Dolon after his capture. By contrasting guarding and
guiding, ships on sea and on land, the transformation of the Homeric
phrase accentuates the greater security brought by Τύχα.32 The
rhythmical departure from Homeric precedent in θοαί / νᾶες high-
lights the thematic differentiations at work.
In choosing an adjective composed of three long syllables
(λαιψηροί), Pindar also seems to be deliberately mobilizing a tension
between sound and sense, but metrical form is also being used to
underscore literary appropriation. Not only is the adjective never
used by Homer to describe war,33 it is never employed by Homer in
this case, and eight of its ten uses in the Iliad are in the neuter plural
with a short final syllable: only once is it used in a form with three
long syllables.34 The adjective usually occurs in cases where there is a
fairly strong coincidence of rhythm and sense, such as Il. 10.358 (γνῶ
ῥ’ ἄνδρας δηΐους, λαιψηρὰ δὲ γούνατ’ ἐνώμα, a line that falls into
spondaic and dactylic halves) and Il. 14.17 (ὀσσόμενον λιγέων ἀνέμων
λαιψηρὰ κέλευθα).35 Pindar’s use of the adjective is therefore un-Homeric
in its semantic application, grammar, and metrical form.36 The word’s
31
This line is repeated at Il. 10.396.
32
Cf. Silk (2007) 184: ‘unlike in epic, the epithet here is also protreptic: ships are
not “swift” unless steered by Tyche’.
33
The adjective is only employed in the Iliad: is Pindar’s use describing ‘war’ here
given added point by his borrowing a word that is only used in Homer’s war poem? In
this respect it may also be significant that νῆες θοαί only occurs in the Iliad.
34
Cf. Il. 21.278 λαιψηροῖς ὀλέεσθαι Ἀπόλλωνος βελέεσσιν.
35
Cf. also Il. 20.93 (εἰρύσαθ’, ὅς μοι ἐπῶρσε μένος λαιψηρά τε γοῦνα); Il. 22.24 (ὣς
Ἀχιλεὺς λαιψηρὰ πόδας καὶ γούνατ’ ἐνώμα).
36
The phrase ἀγοραὶ βουλαφόροι is also Homeric, as noted by Catenacci (2013)
584, and provides another subtle contrast: the only time the phrase is used in epic is at
Od. 9.112, where Odysseus is explaining that the Cyclopes do not have ‘assemblies for
council’ (τοῖσιν δ’ οὔτ’ ἀγοραὶ βουληφόροι οὔτε θέμιστες).
86 Tom Phillips
rhythmical articulation serves less to mirror its meaning than to correlate
with and highlight the complex indirectness of its signification.37
More straightforwardly, the enjambed θοαί / νᾶες may also suggest
the pitching and rolling of a ship in the waves. Such an effect would
reinforce the one already observed by Michael Silk, in which the
rhythm of πόλλ’ ἄνω, τὰ δ’ αὖ κάτω (6, — ˘ — ˘ — ˘ —) also suggests
the image of a ship moving up and down in the water.38 Moreover,
O12s6 as a whole is a fascinating study in the effects Pindar obtains by
juxtaposing dactylic and epitritic metra. The line involves a movement
from epitrite to dactyl and back with long link syllables (E – D – E):
s6:
— — — — — —
˘ ˘ ˘ ˘˘—˘˘———˘———˘—
πόλλ’ ἄνω, τὰ δ’ αὖ κάτω ψεύδη μεταμώνια τάμνοισαι κυλίνδοντ’ ἐλπίδες·
a6:
—
˘———˘———˘˘—˘˘———˘———˘—
ἀντικύρσαντες ζάλαις ἐσλὸν βαθὺ πήματος ἐν μικρῷ πεδάμειψαν χρόνῳ.
s6: men’s hopes roll up and down as they voyage across vain
falsehoods.
a6: having encountered storms they exchange in a short time their
suffering for great good.
In each case, rhythm accentuates meaning. In 6, the short syllables of
μεταμώνια connote movement, while the epitritic rhythm of τάμνοισαι
κυλίνδοντ’ ἐλπίδες echoes the earlier πόλλ’ ἄνω, τὰ δ’ αὖ κάτω, remind-
ing the listener, in conjunction with the verb κυλίνδοντο, of the image
of the ship tossing in the waves. There may also be a subtle contrast
between the long syllables of τάμνοισαι, which may combine with the
verb’s meaning (‘cut’) to indicate decisive movement, and the more
‘up and down’ rhythm of κυλίνδοντ’ ἐλπίδες. In the antistrophe,
37
Silk (2007) 184: ‘it evokes, metonymically, the swift feet of soldiers on the charge,
but also looks ahead to the special “feet” of the victor’. It can also be taken in several
senses: in addition to Silk’s readings see the explanations of the scholia, according to
which λαιψηροί can mean κοῦφοι καὶ ἀνόητοι, on the grounds that men behave
without reason when fighting (διὰ τὸ ἀνοήτως ποιεῖν ὁρμᾶν τοὺς μαχομένους,
Σ O.12.5a = i 350 Dr), or can refer to the speed with which ‘anger’ springs up in
men, resulting in wars (ὀξέως ἡ ὀργὴ διανίσταται τῶν ἀνθρώπων, ἐξ ἧς ὁ πόλεμος,
Σ O.12.5b = i 350–1 Dr).
38
Silk (2007) 185. He also suggests a stylistic enactment of the theme of reversal in
the shift from e – D in s1 to D – e in e1 (p. 179).
however, rhythmical shifts correlate with the notions of exchange and
changes of fortune. While the heavier rhythm of ἀντικύρσαντες ζάλαις
might suggest the weightiness and difficulty of the ‘storms’, the shift
from dactyls to epitritic cola underscores the idea that suffering is
‘exchanged’ for good fortune,39 formally enacting the poem’s themat-
ics of change and development.40
Although far from exhaustive, this survey suggests that the the-
matic connections between words and rhythms were a feature of
Pindar’s style, but it also makes clear the extent to which we need
to grapple with the problems created by our forms of access to the
poems in responding to Pindar’s rhythmical poetics. Quite aside from
the evidential issues outlined above and the inevitable provisionality
of the readings I have offered, the act of writing about such passages,
relying as it does on a process of excerption and analysis, cannot but
(at least to some degree) misrepresent the effects these passages would
have created in the expressive texture of the performance event.41
Imaginative reconstruction requires an awareness of the subtlety and
fleetingness of such effects, as well as the additional affectivity that
would have been created by their mediation by music and dance.
My readings also raise deeper methodological problems. The pas-
sages I have quoted represent only a small fraction of the Pindaric
corpus, and there are many sequences in Pindar where no such
conjunctions are obviously apparent. On this basis, it might be argued
that the passages I have discussed are ‘the most felicitous of acci-
dents’.42 Although this objection neglects the point that the relative
infrequency of such conjunctions would have made them more
noticeable to audiences, it highlights the questions of when and on
what grounds we should detect meaningful intersections of semantics
and form. These questions can be addressed with reference to
remarks by the musicologist Nicholas Cook on the vexed question
of how music might be said to assume or to act as a conduit for
‘meaning’. Cook resists seeing ‘meaning’ either as immanent in the
structures and sounds of music, or as simply arbitrarily imposed on
39
This correlation approximates semantic content (ἐσλὸν βαθύ occupies half of the
dactylic metron) rather than mapping it exactly.
40
For which see e.g. Silk (2007) 194–7.
41
The passages I have examined are also likely to have been some of those in which
the relationship between language and the other elements of the performance was
most straightforward.
42
West (1982) 39, discussing Od. 11.593–600.
88 Tom Phillips
the grounds that because a piece of music does not have propositional
content it can ‘mean’ anything we want it to. Instead, he argues that,
‘regarded as agents of meaning, musical works are unstable aggregates
of potential signification’,43 which have the potential to be under-
stood as bearing meanings as a result of its functions in particular
circumstances.
He illustrates this with the example of the television advert in
which ‘shots of a Citroën ZX 16v powering its way up twisting
country lanes are aligned with extracts from Mozart’s Marriage of
Figaro overture. Heard in this context, the energetic and expressive
attributes of Mozart’s music . . . cluster themselves around the car,
transferring to it the qualities of power and verve and grace associated
with them . . . The music seeks out the qualities of the car, and
conversely the image of the speeding Citroën might be said to inter-
pret the music. And so a composite meaning emerges, one which was
immanent in neither the overture nor the car.’ This example illus-
trates that ‘music [is] never “alone”, that it is always received in a
discursive context, and that it is through the interaction of music and
interpreter, text and context, that meaning is constructed, as a result
of which the meaning attributed to any given material trace will vary
according to the circumstances of its reception’. As a result, ‘it is
wrong to speak of music having particular meanings; rather it has the
potential for specific meanings to emerge under specific circum-
stances’.44 Poetic rhythms are different from the music Cook dis-
cusses, in that they arise in part from dispositions of words, and are
therefore less separable from verbal signifiers than a Mozart melody is
from a piece of film footage. Yet rhythmical structures can usefully be
conceived as ‘potential signfication[s]’, insofar as their salience, sig-
nificance, and form are contingent upon particular conjunctions of
sound and sense. Relations between words and rhythms are unstable
and multiple: a run of short syllables or a sequence of dactyls need not
connote movement or speed, but may be heard in contexts that allow
such associations to emerge.45 But these associations are not arbitary,
in that the meaningfulness of their relation to the properties of the
rhythmical structure is conditioned by those properties and the
perceptions they enable.
43 44
Cook (2001) 188. Cook (2001) 180.
45
Cf. Stewart (2002) 79: ‘sounds in poems are never heard outside an expectation
of meaning’.
The instability and multiplicity of this relationship means that the
mimetic and affective qualities of rhythmical form are likewise vari-
ous. As well as highlighting the expressiveness of Pindar’s poetics, my
readings suggest that the notion of poetic mimesis directed towards
objects in the world which we find in Dionysius and other critics only
gets us so far in conceptualizing the poems’ textures.46 Although there
are numerous passages in which rhythm is straightforwardly suggest-
ive in relation to verbal meaning, rhythmical effects elsewhere do not
so much relate to the properties of things conceived in an abstract or
objective sense as combine with focalizing structures to give shape to
intuitions about the world and momentary experiences of phenom-
ena. The movement of τοὶ μὲν οἰχθεισᾶν πυλᾶν imagines the suspense
felt by the focalizing viewer as well as the physical qualities of the
doors, while Heracles’ wonder-filled gazing at the olive trees in
Olympian 3 correlates form and perception. In these moments, fleet-
ing but intense points of contact are opened up between listeners’
experience, the imaginative worlds of the poems, and the figures who
inhabit them. Elsewhere, in phrases such as θοαί / νᾶες and λαιψηροὶ
πόλεμοι, rhymical distinctiveness accentuates Pindar’s idiomatic re-
configuring of epic language. Common to all of these analyses are
emphases on the poems’ appeals to the senses, and on rhythm as a
vehicle for extra-verbal effects as well as a means of inflecting
meaning.
STANZAIC INTERACTION: PYTHIAN 1
So far I have focused on small-scale effects, but in considering the

description of the Muses and Etna’s eruptions in the opening stanzas
of Pythian 1, I explore how rhythm and meaning can be interrelated
in larger structures, at the level of interaction between stanzas. Criti-
cism of the poem has often noted the connections between Zeus,
Hiero, Apollo, and Pindar himself as guarantors of cosmic, political,
and poetic order,47 and I shall argue that the rhythmical structure of
46
Cf. however Porter (2016) 406–8 on Dionysius’ reading of Od. 11.593–8, who
argues that while Dionysius is working within a mimetic framework, ‘he is more
interested in the art than in the events or their mimesis’.
47
See e.g. Too (1998) 19–22; Athanassaki (2009); Morgan (2015) 308–20.
90 Tom Phillips
the opening triads plays a crucial role in articulating their metapoetic
discourse.
The tradition of remarking on the metrical sophistication of the
Pindaric stanza goes back at least as far as Dionysius of Halicarnassus,
who contrasts the ‘short strophes’ of Sappho and Alcaeus with the
longer periods used by Stesichorus and Pindar, which were ‘divided
into many metra and cola for no other reason that their desire for
variation’ (εἰς πολλὰ μέτρα καὶ κῶλα διένειμαν αὐτὰς οὐκ ἄλλου τινὸς ἢ
τῆς μεταβολῆς ἔρωτι).48 We should not rest content with Dionysius’
explanation, however. The use of more varied metrical structures
opens up the possibility of greater interaction between stanzas than
is possible in shorter strophes, where the sheer number of lines in the
same metrical form makes it difficult for one line to be heard against
or in relation to another across stanzas. In Pindar, by contrast, many
individual metrical lines within a stanza occur only once in a given
poem, allowing listeners to perceive responsions between stanzas, and
in turn to recognize thematically significant connections between
them.49 While greater knowledge of Stesichorus would doubtless
shed further light on the development of this technique,50 Tyrtaeus’
elegies provide a useful comparandum for the kind of effects gener-
ated by interaction between stanzas. It has recently been argued that
Tyrtaeus often employs stanazic responsion for thematic effect, in
order to develop ethical contrasts and articulate his arguments.51
What will emerge from this analysis, however, is that whereas
Tyrtaeus’ ‘stanzaic architecture’ is chiefly a matter of thematic inter-
action at the verbal level, the greater metrical complexity of Pindar’s
poetry allows for the development of complex relations between rhyth-
mical and verbal aspects of the poetry. Concomitant with ‘division into
many metra and cola’ is an increase in rhythm’s semantic agency.52
48
Comp. 19.
49
My arguments operate on the premise that audiences would have been able to
perceive structural connections over short durations: identification of thematic and
rhythmical parallels within triads and beween one triad and the next is likely to have
been facilitated by the chorus’ similar positioning and movement, as well as by their
temporal proximity. Cf. You (1997) 363 on the ordering effects of rhythm, and the
remarks of Prauscello (2013b) 258–9, 270–2.
50
On verbal responsion in earlier poets see Stockert (1969) 26–7. For Stesichorus’
use of dactylic metres see e.g. West (1982) 49–51; Pavese (1997).
51
See especially Faraone (2006) 31–3.
52
Cf. Mullen (1982) 91; Phillips (2013) 55–6.
Pythian 1 is structured in such a way that listeners hear the ecphrasis
of Etna’s eruptions in the second strophe within a rhythmical frame
that has been occupied by the opening description of the Muses and
Apollo, which spans the first strophe and antistrophe. This rhythmical
structure complements the poem’s verbal discourse by enacting song’s
power to order and remake its subjects.53 For convenience, I reproduce
here the relevant stanzas together with metrical analyses:
—
˘———˘———˘˘—˘˘—
sA: Χρυσέα φόρμιγξ, Ἀπόλλωνος καὶ ἰοπλοκάμων 1
—
˘———˘˘——˘——
σύνδικον Μοισᾶν κτέανον· τᾶς ἀκούει
—
˘˘—˘˘———
μὲν βάσις ἀγλαΐας ἀρχά, 2
— — —
˘———˘—
πείθονται δ’ ἀοιδοὶ σάμασιν 3
— —
˘˘—˘˘—˘—˘—
ἁγησιχόρων ὁπόταν προοιμίων
—
˘———˘˘—˘˘—
ἀμβολὰς τεύχῃς ἐλελιζομένα. 4
—
˘———˘———˘—
καὶ τὸν αἰχματὰν κεραυνὸν σβεννύεις 5 5
— — — — — —
˘˘ ˘˘ ˘
αἰενάου πυρός. εὕδει δ’ ἀνὰ σκά-
—
˘˘—˘˘——
πτῳ Διὸς αἰετός, ὠκεῖ-
—
˘˘—˘˘——˘——
αν πτέρυγ’ ἀμφοτέρωθεν χαλάξαις, 6
—
˘———˘———˘˘—˘˘—
aA: ἀρχὸς οἰωνῶν, κελαινῶπιν δ’ ἐπί οἱ νεφέλαν 1
—
˘———˘˘——˘——
ἀγκύλῳ κρατί, γλεφάρων ἁδὺ κλάϊ-
—
˘˘—˘˘———
θρον, κατέχευας· ὁ δὲ κνώσσων 2
— — —
˘———˘—
53
For verbal responsion in the poem see Stockert (1969) 18, 36–7, 53, 60, 69, 84.
92 Tom Phillips
ὑγρὸν νῶτον αἰωρεῖ, τεαῖς 3
— —
˘˘—˘˘———˘—
ῥιπαῖσι κατασχόμενος. καὶ γὰρ βια-
—
˘———˘˘—˘˘—
τὰς Ἄρης, τραχεῖαν ἄνευθε λιπών 4 10
— — — — — — — —
˘ ˘ ˘
ἐγχέων ἀκμάν, ἰαίνει καρδίαν 5
—
˘˘—˘˘——˘——
κώματι, κῆλα δὲ καὶ δαιμόνων θέλ-
—
˘˘—˘˘——
γει φρένας ἀμφί τε Λατοί-
—
˘˘—˘˘——˘——
δα σοφίᾳ βαθυκόλπων τε Μοισᾶν . . . 6
—
˘———˘———˘˘—˘˘—
sB: . . . τᾶς ἐρεύγονται μὲν ἀπλάτου πυρὸς ἁγνόταται 1
— — — —
˘ ˘˘——˘——
ἐκ μυχῶν παγαί· ποταμοὶ δ’ ἁμέραισιν
—
˘˘—˘˘———
μὲν προχέοντι ῥόον καπνοῦ 2
— — — — — — —
˘ ˘
αἴθων’· ἀλλ’ ἐν ὄρφναισιν πέτρας 3
— —
˘˘—˘˘———˘—
φοίνισσα κυλινδομένα φλὸξ ἐς βαθεῖ-
—
˘———˘˘—˘˘—
αν φέρει πόντου πλάκα σὺν πατάγῳ 4
— — — — — — — —
˘ ˘ ˘
κεῖνο δ’ Ἁφαίστοιο κρουνοὺς ἑρπετόν 5 25
—
˘˘—˘˘——˘——
δεινοτάτους ἀναπέμπει· τέρας μὲν
—
˘˘—˘˘——
θαυμάσιον προσιδέσθαι,
—
˘˘—˘˘——˘——
θαῦμα δὲ καὶ παρεόντων ἀκοῦσαι, 6
Golden phorminx, just possession of Apollo and the violet-coiffed Muses,

to which the footstep listens, the beginning of splendour. The singers obey
the signs whenever you strum and strike up the openings of chorus-leading
preludes. You douse the spearing lightning of ever-flowing fire. Zeus’ eagle
sleeps on the sceptre, slackening his wings on both sides,
the lord of birds; you pour a dark cloud over his curved head, a sweet seal
of the eyelids. Slumbering, he raises his rippling back, caught by your
notes. And violent Ares, leaving behind his harsh spear-point, delights his
heart in drowsing, and the darts bewitch the gods’ minds with the skill of
Apollo and the deep-girdled Muses . . .
. . . from whose [sc. Aetna] depths holiest springs of unapproachable fire
belch forth. In the day lava streams pour forth a blazing stream of smoke,
but in the darkness a rolling red flame carries rocks into the deep expanse
of the sea with a crash. That creature sends up most terrible springs of
Hephaestus’ fire, a portent wondrous to behold, and a wonder even to hear
of from those present . . .
The general thematic similarities between the strophe and antis-
trophe of the first triad are clear: both depict the power of music to
calm violent impulses and enchant the mind. Moreover, responsion
reinforces thematic connection between individual elements and
creates interaction between them. One such instance is s3 / a3:
πείθονται δ’ ἀοιδοὶ σάμασιν / ὑγρὸν νῶτον αἰωρεῖ, τεαῖς (— — — ˘ —
——
˘ —). Here, responsion suggests a connection between the eagle’s
movement and that of the singers. Rather than indicating physical
similarity between the referents, the setting of the second phrase in a
rhythmical sequence marked by association with choral dance for-
mally enacts the eagle’s subjection to the music.54 Similar is s5 / a5:
καὶ τὸν αἰχματὰν κεραυνὸν σβεννύεις / ἐγχέων ἀκμάν, ἰαίνει καρδίαν
(— ˘ — — — ˘ — — — ˘ —). Here again rhythmical parallelism under-
scores the similarity of the events they describe.
The verbal and formal congruence of the opening strophe and
antistrophe throws into sharp relief the very different effect created
by the contrasting subject matter of the first two strophes. Thematic
contrasts across stanzas operate at a general level, with Typhos’
activity opposing that of the Muses and Apollo, but also manifest
themselves in numerous pairs of responding lines. The pattern of
rhythmical responsion and semantic contrast begins in s2: μὲν βάσις
ἀγλαΐας ἀρχά / μὲν προχέοντι ῥόον καπνοῦ (— ˘ ˘ — ˘˘ — — —). The
54
Other instances of rhythmical significance: while the rhythm of -ωθεν χαλάξαις
exemplifies a climactic epitritic movement in which heavy syllables accord with the
referent, reflecting the drooping of the wings, the three heavy syllables of ὠκεῖαν
emphasize that the ‘swift’ wings are now at rest.
94 Tom Phillips
former line is significant for setting up the connection between
dactylic runs and the dancers’ movements, which is continued in
line 4 (s4). This association frames the occurrence of the rhythm in
line 22, and stresses the semantic contrast between the glitteringly
ordered movement of the ‘celebration’ and that of the ‘flow of smoke’.
Yet setting the latter line in this rhythm also conveys the sense of
Etna’s activity being transfigured by its realization in poetry, the
rhythmical structure substituting for the amorphous physical reality
of the smoke rising from the lava flows.55
Relations of responsion and contrast continue in s5–6: καὶ τὸν
αἰχματὰν κεραυνὸν σβεννύεις | αἰενάου πυρός. εὕδει . . . / κεῖνο δ’
Ἁφαίστοιο κρουνοὺς ἑρπετόν | δεινοτάτους ἀναπέμπει (— ˘ — — — ˘
———
˘ — | — ˘ ˘ — ˘ ˘ — —). The welling up of fire described in the
second pair of lines contrasts with fire being quenched in the first
strophe, but as in the previous example the rhythmical responsion
forges a connection between the lines that helps to frame the latter’s
meaning with the associations of the former; by recalling the quench-
ing of the thunderbolt, the rhythmical structure enacts the poem’s
containment of the dangerous energies that Typhos unleashes even as
they are brought before listeners’ imaginations.56
Perhaps the most telling of these responsional contrasts is s4. The
line in the second strophe that describes ‘in the darkness a rolling red
flame carries [rocks] into the deep expanse of the sea with a crash’,
expressing the powerful violence of the eruption,57 and vivified by
alliteration and the dactylic rhythm (φοίνισσα κυλινδομένα φλὸξ ἐς
βαθεῖαν φέρει πόντου πλάκα σὺν πατάγῳ, 24), repeats the rhythmical
pattern of the line in the first strophe that describes the ‘chorus-leading
55
The responsion may also make the point that the ῥόον καπνοῦ, both as referent
and sign, is not quite as disorderly and threatening as it might first appear: on this
reading, the stylization of rhythm helps to bring out the order inherent in the referent.
56
Ancient critics were sensitive to the visual power of the opening sequence as a
whole: see e.g. the visually detailed exegesis of πυρὸς ἁγνόταται at Σ P.1.41b (ii 14 Dr),
and the emphasis on the visual elements of the description of the eagle in Σ P.1.10a
and 17b (ii 10, 11 Dr).
57
It is likely that ancient audiences (and readers) will have reacted strongly to the
onomatopoeia of the second strophe in particular. Several of the words used by Pindar
are discussed in Philodemus De poem. 1 in relation to their euphonic effects: with
πόντου πλάκα compare Phld. 1.122.18–19 τοῦ ‘πόντ[ου . . . . . . . . . ] πλάκα’: the source
of the phrase is unclear; see Janko (2000) 335 for discussion. At Phld. 1.93.10–12 the
Homeric phrase ἐρευγομένης ἁλὸς ἔξω (Il. 17.265) is discussed; cf. Pindar’s ἐρεύγονται.
With Pindar’s ἐλελιζομένα compare the discussion of Homer’s uses of ἐλελίζω at
Phld. 1.107.
preludes’ (ἁγησιχόρων ὁπόταν προοιμίων ἀμβολὰς τεύχῃς ἐλελιζομένα, 4).
The metrical structure (— — ˘ ˘ — ˘ ˘ — ˘ — ˘ — — ˘ — — — ˘ ˘ — ˘ ˘ —)
has a different expressive force in each case.58 In the first strophe, the
dactylic movement is co-ordinated with the energetic movement of
the dance, while in the second it suggests the speed with which the
‘flame’ rolls the rocks down to the sea. The sound and rhythm of the
second line are mimetically expressive, but insofar as responsion
invites listeners to hear line 24 as echoing the structure of line 4
and hence attend to the formalizing effects of rhythmical structure,
rhythm accentuates the separation of the text from the phenomena
to which it refers.59
Further examples of such responsions could be cited,60 and
although our appreciation of the poem’s performative qualities is
necessarily diminished by our ignorance of the music and dance,
their general effect is reasonably clear. By pointedly juxtaposing the
Muses and Typhos, Pindar has created a structure that both imitates
and contains the cosmic dissonance Typhos embodies.61 The artful
interlacing of semantic contrast and rhythmical responsion highlights
how the song translates its subject matter into sensuous form, reshap-
ing the dangerous primary qualities of Etna into the stylized beauty of
rhythmically articulated language. Rather than simply ‘weighing
down’ and containing Typhos, as Etna itself does (πιέζει . . . συνέχει,
19), the ecphrasis acts as both a representation of a real-world
phenomenon and the production of an autonomous aesthetic
58
These effects would have been enhanced by the chorus’s gestures and dance
steps, and although we cannot know what this would have entailed, several possibil-
ities suggest themselves. The performance involves a marked gestural shift, with the
chorus performing in a fairly neutral way in the first strophe (imitating choral dance,
and therefore not needing to be especially imitative) and then shifting to more
markedly imitative movements and gestures in the second. Alternatively, the very
absence of such imitation during the description of Etna may have reinforced the
theme of aesthetic transformation by framing the eruptions in the same physical
expressions used in the first strophe. Regardless of the precise nature of the chorus’
movements, their bodily disposition would have acted as a conduit for the imposition
of the framing effect of rhythm and music, and therefore have involved more than
‘mimesis of what is said with gestures’ (Pl. Lg. 816a) in which dance functions as an
extension of vocal utterance. For this notion of dance see Peponi (2009) 59–60.
59
Cf. Porter (2016) 408 on ancient critical engagement with the separation of text
and referent.
60
See e.g. P1s6, where responsion reinforces the thematic similarity between the
verses in the strophe and antistrophe.
61
On the relationship between Etna as a physical environment and the poem’s
configuration of political space see Currie (2012) 297–9; Morgan (2015) 318.
96 Tom Phillips
construct; the very descriptive and affective intensity with which the
text establishes its mimetic connection with its referents is simultan-
eously the means by which it transfigures them.62 As such, the
ecphrasis generates affective force not just through its imagery of
height, overwhelming physical power, and cosmic dissonance,63 but
also through the movement by which it constitutes its own formally
distinct domain decoupled from its real-world entanglements.64 Yet
even as musical and rhythmical structure translates Typhos into an
aesthetic totality and enacts the poem’s imposition of order onto its
materials, the subject matter threatens to overrun the mimetic form
being imposed on it.65
While this dynamic can be read as a formal correlative of the
poem’s simultaneous attempt to transcend the conditions of its emer-
gence and its acknowledgement of the ongoing ethical, political, and
military challenges to which Hiero and his citizens are subject,66 the
musical and verbal totality of which it forms part also has wider
consequences for listeners’ self-conception. The description of Etna
as ‘a wonder even to hear of from those present’ (θαῦμα δὲ καὶ
παρεόντων ἀκοῦσαι, 26) draws on aesthetic vocabulary.67 Like the
scene it describes, the poem is a ‘wonder’ to be admired, but despite
the importance of physicality to the audience’s reaction,68 this ‘won-
der’ entails a form of response that goes beyond the somatic pleasure,
somnolence, and bewitchment created by the Muses’ performance
62
On mimesis as representation and production see e.g. Halliwell (2002) 16–19,
and the discussion of Peponi (2009) 64–5. Cf. also Currie (2012) 296 for Pindar’s
discourse as a response to contemporary volcanic activity.
63
For these features as markers of the sublime in ancient Pindaric criticism see
Porter (2016) 350–60: on the connections between the volcanic imagery at [Long.] De
subl. 35.5 and P.1 see Phillips (2016) 74–5.
64
See e.g. Culler (2015) 34–7, 229 on lyric as aiming at the performative consti-
tution of events.
65
See Fitzgerald (1987) 144–7 on the relationship between liquid imagery, music,
and violence in the opening stanza.
66
On the poem’s relation to its context in this respect see e.g. Athanassaki (2009),
Morgan (2015) 345–6.
67
See e.g. Morrison (2012) 131; Phillips (2016) 150–2. The language of θαῦμα
recalls, for instance, the description of the Delian Maidens at Hom.h.Ap. 156 (πρὸς δὲ
τόδε μέγα θαῦμα, ὅου κλέος οὔποτ’ ὀλεῖται). Cf. also Morgan (2015) 319 on the
connection between Etna as a ‘heavenly column’ (19) and Pindar’s use of monumental
imagery to characterize his poetry.
68
The opening πείθονται δ’ ἀοιδοὶ σάμασιν hints at the impact of the poem on
listeners’ senses: they, like the ‘singers’, will ‘obey’ the music in having their response
to the poem shaped by it.
(ἰαίνει καρδίαν / κώματι, θέλγει φρένας). In order for the tension
between rhythm and language in the second strophe and its structural
relation with the first triad to emerge, listeners are required to
make connections between parts of the poem, attend to the subtle
interaction within the performance of bodily and noetic elements,
and to construct inferences about the wider significance of structure.
Rather than being passive recipients of bodily affect in the manner
of Typhos and the Muses’ audience, Pindar’s listeners are simultan-
eously subjects of enchantment and interpreters of form.
CONCLUSIONS
This chapter has focused on the interrelations between words and

their rhythmical frames. In part, this method is dictated by the
absence of definite knowledge of dance moves and musical accom-
paniment. Whatever these elements consisted of, however, the rela-
tionship between words and rhythm would have been a constituent
feature upon which performers would have drawn, and can therefore
offer valuable clues about what performances would have looked,
sounded, and felt like, even if our approximations of this are neces-
sarily limited. Yet even seen on their own terms, the relations
between words and rhythms have a great expressive and interpret-
ative richness. The above readings have highlighted the varieties of
rhythmical enactment at work in Pindar’s poetry, ranging from
relatively simple conjunctions of form and semantic content, to the
more subtle effects created when rhythms are employed to accentu-
ate an image or to encourage a particular way of understanding a
word or phrase.
Although Pindar’s techniques have long been recognized as a
development from the simpler structures seen in earlier poets such
as Alcman and Stesichorus,69 less attention has been paid to the role
of metrical complexity in opening up possibilities for the interaction
of verbal and rhythmical forms. Greater intrastanzaic variety allows
for the development of intra- and intertriadic relations between
individual lines, which in turn allows for interaction between these
69
See e.g. Mullen (1982) 91; West (1982) 60; Davies and Finglass (2014) 47–52.
98 Tom Phillips
lines that adds to their significance. The ‘love of variety’ attributed to
Pindar and his contemporaries by Dionysius has numerous conse-
quences for how the poems construct and project meaning. Use of
authors such as Aristides Quintilianus and Dionysius of Halicarnas-
sus as guides to the affectivity of choral poetry needs to be supple-
mented, I have argued, by close attention to the variousness with
which rhythms and words interact. Perhaps most importantly, I hope
to have recovered something of Pindar’s extraordinary sophistication
as a musical as well as a verbal artist, and to have demonstrated
that attention to this aspect of his texts is not only interpretatively
rewarding, but can also bring us a little closer to the τέρψις of his
early listeners.
Music in Euripides’ Medea

Oliver Thomas
Among the major developments in the study of Greek tragedy of the

last forty years has been the increasing tendency to regard our texts as
the written residue of a multifaceted performance, where they com-
bined with acting, movement, etc. to give more than the sum of the
parts. Within this trend, however, scholars have made only limited
progress in analysing the contribution of music—naturally enough
given our scanty sources.
In several plays, the distribution of musical versus spoken sections
appears deliberate. For example, in his Electra Euripides saved up the
‘recitative’ sound of anapaests for the entrance of Clytemnestra (line
988), and contrasted the sparse use of music in lines 1–1146 (c.22 per
cent of lines accompanied) with the dénouement as the characters
respond to Clytemnestra’s assassination (lines 1147–1359, c.71 per
cent accompanied); the soundworld shifts with the plot. More ambi-
tious approaches have tried to find patterning of rhythms within a
play’s musical sections, or to analyse the politics of which characters
are given song.1
Perhaps the most stimulating insight into Euripides’ music has
come from contextualizing it within the increasing musical profes-
sionalization and change of c.440–380—the so-called ‘New Music’.
Of the composers associated with these developments in antiquity,
Euripides is one of the most senior, and Csapo (1999–2000) has
I am grateful to Oliver Taplin, Pauline LeVen, Patrick Finglass, David Creese, Judith
Mossman, and audiences in Southampton, Oxford, and Newcastle for their advice.
1
See e.g. Scott (1984); (1996); Chiasson (1988) for rhythmic patterning; Hall (1999)
for singing and social status.
100 Oliver Thomas

argued persuasively that he was no reactive dabbler in an elitist avant-
garde. Rather, the new styles were popular (though easily pilloried),
and Euripides’ experimentation with them can be dated before 420,
i.e. before much of the evidence for developments in dithyramb.
Csapo studies in particular Euripides’ use of actors’ monodies and
duets, trochaic recitative, astrophic choruses, and the tendency for
these forms to be assigned to female or non-Greek characters in
emotional situations. These practices distinguish Euripides from
Sophocles, so far as we can tell.
In this paper I shall argue that Euripides already in 431 was
engaged with the discourse and possibly the practice of musical
novelty, in a different way from the features discussed by Csapo.2
The second half of the paper will discuss what characters within the
Medea say about music. Before that, I shall discuss and defend the
plausibility of a difficult testimonium in Athenaeus, which appears to
say that the play was innovative in its treatment of melody.3 These
two issues have been studied separately before, but connecting them
will enhance our appreciation of both: the passages of Medea will
support Athenaeus’ testimony, at least to the extent that they chime
remarkably well with it; and Athenaeus’ testimony, if correct, affects
the way in which the passages of Medea should be interpreted.
ATHENAEUS’ TESTIMONY
The testimonium for the Medea’s musical practice is embedded in two

passages of Athenaeus, where characters discuss an intriguing dramatic
work by Callias of Athens (7 276a, 10 453c–454a; cf. 10 448b).4
2
Contrast also Weiss’s demonstration (this volume) that throughout his extant
works Euripides played with the instrumental possibilities where his auloi-
accompanied chorus discusses a mythological syrinx-accompanied dance.
3
The evidence for tragic melody, apart from Dionysius of Halicarnassus’ discus-
sion of the chorus’s opening words in Orestes (De comp. 11 = DAGM no. 2), and the
Orestes Papyrus (DAGM no. 3), is scattered and potentially distorted by the system-
atizations of harmonic theory from Eratocles on (see e.g. West (1992b) 184–5).
4
Significant discussions include Pöhlmann (1971); Rosen (1999); Ruijgh (2001);
Smith (2003); Gagné (2013). Koller (1956), D’Angour (2006a) 281–2, and Phillips
(2015) accept them—more easily than I do—as evidence for Euripides’ melodic
practice. Phillips considers its possible ramifications for the portrayal of Echo in the
Andromeda of 412.
Music in Euripides’ Medea 101

At the start of Book 7 Cynulcus mentions the number of potential
puzzles in Clearchus of Soloi’s work On Brainteasers (griphoi), and
gives an example:
εἰ δ’ ἀπιστεῖς, ὦ ἑταῖρε, καὶ τὸ βιβλίον κεκτημένος οὐ φθονήσω σοι, ἀφ’ οὗ
πολλὰ ἐκμαθὼν εὐπορήσεις προβλημάτων. καὶ γὰρ Καλλίαν ἱστορεῖ τὸν
Ἀθηναῖον γραμματικὴν συνθεῖναι τραγῳδίαν, ἀφ’ ἧς ποιῆσαι τὰ μέλη καὶ
τὴν διάθεσιν Εὐριπίδην ἐν Μηδείᾳ καὶ Σοφοκλέα τὸν Οἰδίπουν.
If you don’t believe me, my friend, I do own the book and will happily
lend it to you. You will learn much from it, and have a good store of
posers. For instance, he records that Callias of Athens composed a
tragedy which was grammatike, and that it was on the basis of it that
Euripides wrote the melodies and delivery[?] in the Medea, and that
Sophocles wrote his Oedipus.5
This information is indeed puzzling, as Cynulcus says. Is this the
famous Athenian writer Callias? If so, what is a comedian doing
writing a tragedy? Is grammatike a predicative adjective (‘a tragedy
which was lettered’) or an object complement (‘a tragedy Learning
One’s Letters’), and what would either entail? Assuming from these
two puzzles that it was not a normal tragedy, could it have been a
major influence on two such canonical tragedies as Medea and
Oedipus (Tyrannus, as specified later)?
Athenaeus raises these questions only to abandon them immedi-
ately: he sets us a brainteaser, then suspends the solution until the
middle of Book 10.6 There Aemilianus Maurus, in moving the discus-
sion on from drinking-culture, proposes an enquiry about sympotic
brainteasers, ‘not in the manner of the work entitled Lettered Tragedy
by Callias of Athens’ but proceeding from a definition of brainteasers
to comic discussions of them and eventually to the forfeits for failing
to find a solution (448b).7 The host Larensi(u)s responds with a dis-
quisition which starts from and repeatedly returns to Clearchus’ On
Brainteasers, which is clearly Athenaeus’ main source.8 In particular,
5
I discuss below the construction of grammatike and whether διάθεσις does mean
‘delivery’. For Clearchus see Wehrli (1948); Tsitsiridis (2013).
6
Similarly Smith (2003) 315.
7
Aemilianus’ wording τὴν . . . ἐπιγραφομένην γραμματικὴν τραγῳδίαν still leaves the
work’s title ambiguous, in a way I cannot translate. As with Cynulcus’ wording (above),
it could also mean ‘the tragedy entitled Learning One’s Letters’. But it will presently
become clear that the play was not a ‘tragedy’, so that word must be part of the title.
8
448c–e, 452c, 452f, 454f, 455b, 457c (followed by a lengthy paraphrase from
Clearchus’ On Proverbs). Athenaeus is also our main source for Clearchus’ Περὶ
102 Oliver Thomas

Larensius’ information on Callias overlaps with what Cynulcus expli-
citly drew from Clearchus—notably the claim that he influenced both
Medea and Oedipus Tyrannus. Clearchus distinguished seven types of
brainteaser (448c). Larensius does not list them all, but he subsequently
mentions that in On Proverbs Clearchus made one category of brain-
teaser ‘things pertaining to the study [θεωρία] of letters and syllables’
(457e). To this would belong Castorion’s Hymn to Pan (composed of
interchangeable iambic metra containing eleven letters each) and
Lasus’ asigmatic hymns, which Larensius cites just after Callias (454f,
455c). Probably Callias’ play, with its versified alphabets and syllabar-
ies, and characters describing letter-shapes, fell into this category.9
Interestingly, Larensius cites an alternative title Γραμματικὴ Θεωρία,
Studying Letters (453c).10 This title may explain Maurus’ implication
that Callias conducted an unmethodical ‘enquiry’ into griphoi.
Larensius goes into more detail than Cynulcus. Callias’ prologue
involved a character reciting the Ionic alphabet, and was followed by a
chorus based around the chanting of consonant-vowel syllables
which characterized primary education in the larger centres:11
ὁ χορὸς δὲ γυναικῶν ἐκ τῶν σύνδυο πεποιημένος αὐτῷ ἐστιν ἔμμετρος ἅμα
καὶ μεμελοποιημένος τόνδε τὸν τρόπον·
βῆτ’ ἄλφα βα, βῆτ’ εἶ βε, βῆτ’ ἦτα βη,
βῆτ’ ἰῶτα βι, βῆτ’
οὖ βο, βῆτ’ ὖ βυ, βῆτ’ ὦ βω,12
φιλίας, Γεργίθιος, Ἐρωτικά, Περὶ βίων, and Περὶ ἐνύδρων, and a major source for
his Παροιμίαι.
9
I believe one can safely infer that the fragment at 454a, where a woman describes
the shapes ΨΩ as the start of a name she is ashamed to be ‘swollen’ or ‘pregnant’ with
(cf. ψώα ‘stench’), is also from the Lettered Tragedy. As Slater (2002) 127 notes, ΨΩ is
thematic in that it also featured as the last syllable of the chorus discussed below; he
also observes that Ψ and Ω, as newcomers to the Attic alphabet, are aptly figured as a
bastard child.
10
None of the examples in LSJ s.v. θεωρία III.3 mean ‘spectacle’, and I doubt that
sense here. This title might itself pun on the sense ‘sacred embassy’ (in that the
performers have come to offer their play to Dionysus), but there is no sign that the
plot involved an embassy.
11
Syllabaries in antiquity: Rix (1991) Cr 9.1 (Etruscan, c.650); Pl. Plt. 277e–278c;
Crat. 424bc; Quint. 1.1.26–31; Morgan (1998) 59; Cribiore (2001) 172–3.
12
The colometry is speculative. Consonant-names are either trochees or long
monosyllables; metrical responsion requires elided trochees and monosyllables followed
by hiatus. ζξψ appear not to have closed the preceding syllable (cf. West (1982) 17). The
distribution of short syllables suggests iambic or trochaic metre. This suggests βᾱ, βῑ, βῠ,
to give — — ˘ — — — ˘ — — ˘ — — ˘ — ˘ — — — ˘ — — ˘ — — —. Given the rarity of ˘ — ˘ in

καὶ πάλιν ἐν ἀντιστρόφῳ τοῦ μέλους καὶ τοῦ μέτρου γάμμα ἄλφα, γάμμα
εἶ, γάμμα ἦτα, γάμμα ἰῶτα, γάμμα οὖ, γάμμα ὖ, γάμμα ὦ, καὶ ἐπὶ
τῶν λοιπῶν συλλαβῶν ὁμοίως ἑκάστων· τό τε μέτρον καὶ τὸ μέλος ἐν
ἀντιστρόφοις ἔχουσι πᾶσαι ταὐτόν, ὥστε τὸν Εὐριπίδην μὴ μόνον
ὑπονοεῖσθαι τὴν Μήδειαν ἐντεῦθεν πεποιηκέναι πᾶσαν, ἀλλὰ καὶ τὸ
μέλος αὐτὸ μετενηνοχότα φανερὸν εἶναι. (453de)
As for the chorus of women, he has composed it metrically from paired
letters, together with the following type of melody:
Bet’-alpha ba, bet’-e be, bet’-eta bē,
Bet’-iota bi, bet’-
o bo, bet’-u bu, bet’-ō bō.
And again in a responsion [antistrophos] of tune and metre, gamma-
alpha, gamma-e, gamma-eta, gamma-iota, gamma-o, gamma-u,
gamma-ō, and so on for each remaining set of syllables: they all have
the same metre and melody in responding stanzas [antistrophoi]. Hence
Euripides not only is suspected of having composed his entire Medea
from that source, but has patently borrowed the tune itself.13
Larensius specifies that what influenced Euripides was Callias’ open-
ing chorus, whose melody Athenaeus presupposes his readers will be
able to supply from their schooldays. Larensius emphasizes that the
melody repeated in each stanza, hence disregarding the natural
accentuation of consonant-names (e.g. βῆτα vs κάππα) and the nor-
mal avoidance of hiatus (to allow e.g. μῦ εἶ με).14 Such simplifications
do seem plausible for a classroom and mnemonic context. Callias’
chorus caused a (mere) suspicion that ‘the Medea in its entirety’ was
based on it, while Euripides ‘patently borrowed the tune itself ’.
There follows a comment about Callias’ influence on Sophocles.
‘And people say that Sophocles, after hearing this, took the license of
dividing his poetry by its metre[?] (διελεῖν . . . τὸ ποίημα τῷ μέτρῳ).’
This is exemplified, in the manuscripts, by a deformation of OT
332–3. The comment is shown to be parenthetical by the reversion
trochees until Euripides’ Helen (West (1987) 52–5), this is best interpreted as eight
iambic metra. There was probably at least one licentious elision (see below, n. 15), the
best place for which seems to be after — ˘ — ˘ — — (Dale (1968) 72). The reconstruction
by Ruijgh (2001) 260–1, 293–8 is inadmissible, since his view that our text of Athenaeus
is an abbreviation, to which one can liberally restore small words, has been superseded
(Letrouit (1991); Rodríguez-Noriega Guillén (2000)).
13
The first use of ἀντίστροφος with a genitive is unusual. For ἐν ἀντιστρόφοις cf.
[Arist.] Prob. 19.15 918b13. μεταφέρω is either ‘transfer’ or ‘modify’, but
recognizably—hence ‘borrowed’.
14
Hiatus is also admitted around the letter-names in Callias’ prologue (453d).
104 Oliver Thomas

in the next sentence to the topic of antistrophes, which is not relevant
to Sophocles’ passage since that is spoken.15
διόπερ οἱ λοιποὶ τὰς ἀντιστρόφους ἀπὸ τούτου παρεδέχοντο πάντες, ὡς
ἔοικεν, εἰς τὰς τραγῳδίας.
Hence later people, apparently, all borrowed antistrophes into their
tragedies from this source.16
Thus Callias’ use of antistrophes is the key point on both sides of his
alleged influence on Euripides. Clearchus evidently knew that met-
rically corresponding antistrophes occurred in earlier tragedy, so the
emphasis is on Callias’ repetition of melody as well as metre, which
I shall call ‘melodic responsion’. The emphasis implies that melodic
responsion features within the claims which are crucial for us—(i)
that Euripides clearly borrowed τὸ μέλος αὐτό from Callias’ chorus,
and (ii) that there was reason to suspect that the influence of that
chorus extended throughout the whole Medea; Cynulcus’ counterpart
claim spoke of the pervasive influence on Euripides’ μέλη καὶ διάθεσιν.
Where exactly is melodic responsion relevant here? If we make it
part of claim (i), we need to attribute to μέλος a sense like ‘technique
of melodization’, and to envisage some second influence which Cal-
lias might have exerted on Medea as a whole—perhaps covered by
Cynulcus’ word διάθεσις, but quite unclear in its nature. More plaus-
ibly, melodic responsion is part of claim (ii), μέλος in claim (i) simply
means ‘tune’, and διάθεσις (‘organization, delivery’) is a suitably
cryptic way for Cynulcus to allude to melodic responsion.17
15
Compare Larensius’ aside on Stesichorus at 451d. Sophocles’ lines do contain an
ἀντίστροφος in the sense ‘crasis’ (Σ Ar. Pl. 3), but Larensius can hardly switch
meanings so abruptly. Perhaps Clearchus explained the point about Sophocles clearly.
I would restore the quotation from OT: the manuscripts’ deformation, ‘I shall cause
grief neither to myself nor to you if convicted of this’, is plausibly an attempt to suit
meaning (not form) to context, in that it could represent Sophocles asking for
leniency. διελεῖν refers to word-divisions at 453c, f, and OT 332 ταῦτ’ is (probably
along with OC 1164 μολόντ’) the most striking instance in Greek drama of elision at
verse-end. I infer that Callias’ chorus exhibited a similar anomalous elision. Similarly
Ruijgh (2001) 315–18. Smith (2003) suggests very differently that Clearchus’ point
was the appearance of letter-names within tragic language (e.g. ταῦ in ταῦτ᾽), but why
would anyone have chosen OT 332–3 to exemplify this pervasive feature of language?
I do not understand how Smith (2003) 326 takes διελεῖν . . . τὸ ποίημα τῷ μέτρῳ and
its connection to the quotation.
16
I have taken τούτου like τοῦτ’ in the previous sentence; it might also refer to
Callias or Euripides.
17
D’Angour (2006a) 276–82 does take μέλος as ‘technique of melodization’. Hense
(1876) understood διάθεσις to refer to speaker-divisions within the fifth stasimon, but

Interim conclusions: (i) Larensius says that Euripides clearly bor-
rowed Callias’ tune; (ii) Clearchus said that Euripides perhaps bor-
rowed Callias’ practice of melodic responsion throughout Medea.
Corollary: Clearchus supposed that previous antistrophic stanzas
corresponded in metre and rhythm (and probably tempo and mode
and perhaps choreography) but were melodized in some other way,
probably with stricter regard for accent-patterns. We have no other
evidence about how (for example) Pindar and Aeschylus set their
antistrophes.18 Composing a tune which both followed accents and
was recognizably the same in strophe and antistrophe would be
restrictive, even if an audience’s perception of ‘sameness’ allowed
slight variations (as with metre).19
So far we have been concerned with interpreting what Athenaeus
and Clearchus meant. But were their ideas about earlier musical
practice correct? I shall argue that something valuable can be
extracted from them, whereas some scholars have interpreted claims
(i) and (ii) as the worthless result of catastrophic misunderstanding.
I think two main ideas have motivated this belief. The first is that
Larensius dates Callias ‘shortly before Strattis’ (fl. c.400), which is
imprecise if he was active before 431, especially given the more
obvious chronological landmark of Aristophanes. This has led to
three further arguments for distinguishing a second, otherwise
unknown Callias: Athenaeus consistently speaks of ‘Callias of Athens’
when talking about the Lettered Tragedy but of ‘Callias’ when talking
about the famous comedian; the play’s use of the Ionic alphabet
they do not obtain in ‘all’ the Medea. Ruijgh (2001) 273–4 took διάθεσις as the general
structure of prologue, chorus, episodes alternating with stasima etc., but Medea is not
especially innovative in this respect. For ‘delivery’ see LSJ s.v. διάθεσις I.2b.
18
Dale (1968) 204–6 admits our ignorance but inclines to think that antistrophic
music normally included melodic reponsion; similarly West (1992b) 209–12. Koller
(1956) believes Clearchus, but the alleged support from Plato and Aristoxenus (pp. 23,
28) is based on misreadings. For principles by which several later composers correlate
pitch and accent see Cosgrove and Meyer (2006); cf. Winnington-Ingram (1955)
64–73.
19
With melodic responsion, the principle that an accented syllable bears the
(equal-)highest note in its word would often constrain two syllables to the same
pitch, e.g. the second and third syllables of A. Pers. 65 πεπέρακεν ~ 73 πολυάνδρου.
Other tendencies, such as the avoidance of a rise in pitch during a circumflex and of a
fall in pitch during an acute, would constrain e.g. Pers. 67 γείτονα ~ 75 θεῖον ἐ-.
D’Angour (2013) 206–8 conjectures that early composers accommodated accents to a
‘repeated’ melody by admitting a few swaps in the sequence of pitches; cf. D’Angour
(this volume) at n. 15.
106 Oliver Thomas

points to a date around 403; and the Lettered Tragedy is not in the
Suda entry for the famous Callias (κ.213). However, none of these
arguments is weighty. Positing a second Callias does not stop the
reference to Strattis being odd, since only here is Strattis used as a
chronological landmark. The insistence on Callias being ‘of Athens’
here can be ascribed to Clearchus’ influence; when Athenaeus cites
Clearchus for ‘Epaminondas the Theban’ (13 590b), there is no
question of postulating an otherwise unknown Epaminondas.
Inscriptions demonstrate that the Ionic alphabet was widely known
in Athens from c.450. And we do not know how soon after Clearchus
the Lettered Tragedy was lost, or the origins of the Suda’s information.
These are slender grounds for cloning a comedian.20
The second factor which seems to underlie scholars’ scepticism is a
sense that Callias influencing Euripides is preposterous. Here, how-
ever, one must avoid discarding the baby with the bathwater. Both
our central claims involve saying propter hoc where the evidence
almost certainly admitted only post hoc. But such a slide is not
uncommon in the Greeks’ obsessive imposition of an inventor-and-
influence teleology on the history of music.21 Furthermore there is
evidence that all three of Clearchus, Larensius, and Athenaeus are
using the topic of brainteasers to set puzzles for their respective
audiences.22 This playful mise-en-abyme means we should expect
that some claims will appear in a form more paradoxical than was
warranted by the sources on which they were based. Hence a viable
20
If Callias test. 4 is indeed about him (as accepted in K-A and Millis and Olson
(2012) 225–7), it gives two further titles not mentioned in the Suda. For more detailed
summaries of either side of the debate outlined in this paragraph, see Ruijgh (2001)
269–71; Gagné (2013) 304 n. 21. An alternative rationale for the mention of Strattis
will be mentioned below.
21
For this way of thinking see Barker (2014), particularly on Heraclides and
Aristoxenus.
22
For Athenaeus see p. 101 on Cynulcus’ mention of Callias. For Clearchus see
p. 108 on his treatment of Callias’ prologue. Larensius makes a puzzle out of his three
references to forfeits for failing to solve a puzzle, though editors have not understood
this. At 10.448e Larensius describes the forfeit vaguely (‘they used to drink the cup’)
and challenges Ulpian to make sense of it. At 457c he specifies that the cup was
diluted, and again poses a question based on his paradoxical phrasing (‘What pun-
ishment was undergone . . . if in fact they used to drink a diluted cup?’); Dobree’s
tentative <ἅλμῃ> κεκερασμένην, ‘diluted <with brine>’ (1833, 329), accepted by Kaibel
(1923–5) and Olson (2006–12), produces a question which non-sensically answers
itself. In fact, only at 458f–459a does Larensius finally (ἤδη) reveal the solution to the
paradox—that the cup was diluted with brine and had to be downed.

model of claim (ii), for example, is that Clearchus believed that
Euripides in 431 was the first tragedian to use melodic responsion,
and suggested (wrongly but characteristically) that he got it from
Callias, who had used it earlier under special circumstances. One
need not discard the basic belief along with the dubious suggestion.
As for claim (i) about ‘the tune itself ’, I see no reason to rule out a priori
the idea that Euripides did adapt a schoolroom chant, doubtless for a
shocking and significant effect.23 Unfortunately, we are not in a position
to narrow down where in the Medea such an effect is most likely to have
been deployed. I shall therefore leave this issue aside and focus in what
follows on the more tractable question of melodic responsion.
The grounds for doubting Clearchus’ testimony thus exist but are
not forcing. Even if granted, such scepticism can be fleshed out into
more or less plausible models. To take an example, Pöhlmann (1971)
239 and Rosen (1999) argued that Callias was an unknown later
comedian of c.400, who made an absurd joke about how a quotable
(hence earlier) play by Euripides had ‘copied’ (hence followed) the
musical tactics of the present performance. This model not only entails
a dubious second Callias, but also either that Clearchus knowingly
promulgated an absurdity, or that he was confused by an obvious
joke in a source with which he engaged very carefully. This engagement
is clear especially from his treatment of Callias’ prologue, where the
speaker named the twenty-four letters, probably represented as the
twenty-four choreuts as they are introduced to the audience. Ruijgh
(2001) 286, 289 reports the principal manuscript at this point:
πρόλογος μὲν αὐτῆς ἐστιν ἐκ τῶν στοιχείων, ὃν χρὴ λέγειν ἐκ τῶν
στοιχείων διαιροῦντα κατὰ τὰς πάσας γραφὰς καὶ τὴν τελευτὴν κατασ-
τροφικῶς ποιουμένους εἰς τ̅ά̅λ̅φ̅α̅, β̅ῆ̅τ̅α̅, γ̅ά̅μ̅μ̅α̅, δ̅έ̅λ̅τ̅α̅, ε̅ἶ̅τ̅α̅, θ̅ῆ̅τ̅α̅, θεοῦ
γὰρ εἰ γε ἰ̅ῶ̅τ̅α̅, κ̅ά̅π̅π̅α̅, λ̅ά̅β̅δ̅α̅, μ̅υ̅, ν̅υ̅ ξ̅ε̅ι̅ το ο̅υ̅ π̅ε̅ι̅ ρ̅ω̅ σ̅ι̅γ̅μ̅α̅ τ̅α̅υ̅ υ̅ παρον
φ̅ε̅ι̅ χ̅ε̅ι̅ τε τω ψ̅ εἰς το ω̅.24
23
A roughly comparable manoeuvre could be Mahler’s inclusion of the children’s
round ‘Bruder Martin’ in the third movement of his First Symphony. This is uncon-
ventional (and flagged as such by the instrumentation) and demands interpretation:
see e.g. Roman (1973); Jung-Kaiser (1997) 115–25. An alternative interpretative move
in our case is to think that Larensius/Athenaeus is emboldening a more nuanced
statement in Clearchus.
24
Ath. 10 453cd. The sense probably began ‘Its prologue is composed from letters.
People must recite it by dividing it into words according to the side-markings, and by
making it end by reverting to alpha.’ Then there is a quotation largely composed of the
alphabet in order, with several difficulties of detail.
108 Oliver Thomas

This passage is clearly corrupt, for example in its accents and treat-
ment of the sequence εἶ ζῆτα ἦτα. In particular, πάσας γραφάς must be
emended via 453f: ‘Which one must divide in reading according to
the παραγραφάς, just as before.’25 The manuscripts’ markings above
the letter-names are a reasonable guess at what these ‘marks beside’
looked like in Clearchus. They indicate how to ‘divide’ the text, i.e. to
extract words from unspaced majuscule. If the letter-names were
originally written out, as in the manuscript, the reader would not
need these paragraphai, since the unproblematic sequence . . .
ΑΛΦΑΒΗΤΑΓΑΜΜΑ . . . would clarify even the pitfalls of . . . ΝΥΞ-
ΕΙΤΟΟΥΠΕΙΡΩ . . . . Rather, the manuscript’s ψ (not ψεῖ) is the last
trace that in Clearchus the reader was faced with mouthfuls like . . .
Κ̅Λ̅Μ̅Ν̅Ξ̅ΤΟΟ̅Π̅Ρ̅ . . . from which one had to extract trimeters by
reading the marked letters as letter-names. Clearchus thus converted
Callias’ prologue into a riddle for his readers.26 Such engagement
makes it difficult to believe that Clearchus misunderstood a glaring
anachronistic joke.27 Furthermore, creating textualized riddles from
source-material is quite different from promulgating a potentially
confusing joke, and Clearchus’ high-minded comments on the close-
ness of brainteasers and philosophy (457c) make it unlikely to me that
he perpetrated the latter.
Welcker (1832) 152 already hinted at a more attractive explanation
for how Clearchus could have radically misjudged Euripides’ music.
I mentioned that Larensius’ dating of Callias via Strattis calls for an
explanation. One might infer that Clearchus originally cited Strattis
before or à propos Callias. Then one possibility is that Strattis in
his Medea lampooned Euripides by saying something like ‘Do you
remember Callias’ show with that repetitive chorus? That’s where
Euripides got his tunes from for Medea.’ Such a comment, in a
comedy, need not have picked on Medea because it did something
musically remarkable. (Equally, however, such an innovation could
25
Ruijgh (2001) 287–8 accepts πάσας γραφάς as ‘complete letter-names’, without
giving a parallel; in his treatment of 453f he bizarrely suggests κατὰ τὰς <πάσας
γραφὰς> alongside <κατὰ τὰς> παραγραφάς (ibid. 319–20).
26
Clearchus’ intervention here is emphasized rightly by Smith (2003) 318–20.
Athenaeus duplicates it: his audience, like Clearchus’, face a reading puzzle; Larensius’
(aural) audience, like that of Callias, have that puzzle resolved for them.
27
Clearchus also had ready access to information about dramatic dates if he
wanted to check, since his relationship with Aristotle is considered to have extended
into the latter’s mature period (e.g. Tsitsiridis (2013) 4–5).

have been precisely Strattis’ point, or Clearchus might have men-
tioned Strattis for an unconnected reason.)
To summarize, there are models which allow one to discredit
Athenaeus’ testimony. They require one, I think, to hypothesize
fragments of Old Comedy and to attribute a dim reading of them to
Clearchus, whose remains (largely mediated by deipnosophists,
admittedly) are a lively read with little evidence of sloppiness.28
Such models cannot be falsified. However, I have argued that there
is also a viable model which allows (with external justification) that
the claims in the testimonium has been affected by post hoc ergo
propter hoc thinking, and that their surprise-value may have been
increased in transmission, and leaves a late fourth-century scholar
with some reasons for believing that Euripides’ Medea was the first
tragedy to use melodic responsion.
Faced with these alternative approaches to Athenaeus, let us turn
to see whether anything in the text of the Medea favours one over
the other.
MUSICAL DISCOURSE IN MEDEA
The first passage which raises the theme of musical novelty is the
nurse’s parting comment at Med. 190–203. The chorus have arrived
at the sound of Medea’s indoor cries, and suggested that the nurse
should fetch her for a therapeutic chat.29 The nurse doubts that
Medea will accept, since she is refusing advice (184–9), and adds:
σκαιοὺς δὲ λέγων κοὐδέν τι σοφοὺς 190
τοὺς πρόσθε βροτοὺς οὐκ ἂν ἁμάρτοις,
οἵτινες ὕμνους ἐπὶ μὲν θαλίαις
ἐπί τ’ εἰλαπίναις καὶ παρὰ δείπνοις
ηὕροντο βίῳ τερπνὰς ἀκοάς·
στυγίους δὲ βροτῶν οὐδεὶς λύπας 195
28
On scientific matters, however, see Plu. Fac. Lun. 920e πολλὰ τοῦ Περιπάτου
παρέτρεψεν ‘He subverted many Peripatetic ideas’ (fr. 97 Wehrli).
29
In 175 the proposed conversation is metaphorically called an ὀμφά (roughly, a
divine prophetic voice). The chorus may thereby cast their advice in the terms of
another form of music, namely sung hexameter prophecies; the metaphor would
contrast with Aegeus’ actual advice from Delphi later in the play.
110 Oliver Thomas
ηὕρετο μούσῃ καὶ πολυχόρδοις

ᾠδαῖς παύειν, ἐξ ὧν θάνατοι
δειναί τε τύχαι σφάλλουσι δόμους.
καίτοι τάδε μὲν κέρδος ἀκεῖσθαι
μολπαῖσι βροτούς· ἵνα δ’ εὔδειπνοι 200
δαῖτες, τί μάτην τείνουσι βοήν;
τὸ παρὸν γὰρ ἔχει τέρψιν ἀφ᾽ αὑτοῦ
δαιτὸς πλήρωμα βροτοῖσιν.
You’d not be wrong to call former mortals maladroit and not clever at all,
since they invented songs at festivities, at banquets, and alongside dinner-
parties as sounds bringing pleasure to our life, whereas no mortal discovered
how to stop Stygian pain with the Muse and with songs on many strings—
pains from which deaths and terrible misfortunes overturn households. Yet
these are what it profits mortals to cure with singing. Why do they vainly
strain their cries where the dining is fine? After all, the satisfaction of the
dinner is already present and intrinsically brings mortals pleasure.
The nurse dismisses former musical inventions (194 ηὕροντο, 196
ηὕρετο) which accompany commensality, and implicitly calls for a
new, analgesic style.30
Old and new here can take their bearings from two possible deictic
centres—the dates of performance and of action. There are two
reasons to accept the level on which the nurse’s comment is a
metapoetic provocation by Euripides to his audience. First, she dis-
cusses singing rather than spoken conversation, and this points to the
dramatic stylization by which the chorus sings its part. Secondly, she
addresses a vague masculine second person (190 λέγων), rather than the
female chorus, and this would have helped the original audience posi-
tion themselves as addressees. Pucci (1980) 24–32 pressed this meta-
poetic approach furthest, and saw in this passage Euripides’ proto-
Aristotelian definition of tragedy as the genre of psychological catharsis.
However, tragic metapoetics do not steamroller characterization.31
The comments unavoidably come across to the audience also as those
30
The nurse’s three words for commensality cannot be sharply distinguished, but
combine to cover a full range of contexts from public festivals to formal dinners (e.g.
for a marriage) to smaller private affairs. Vox (2003) 831–2 compares Stesichorus fr.
172 Finglass-Davies. The nurse therefore targets a wide range of genres (ὕμνοι in 192
need not mean ‘hymns’ specifically).
31
Compare Torrance (2013) 268: ‘the metapoetic strategies used allow for two
levels of meaning, making sense within the fiction but also serving as markers of
artificiality’.

of a nurse in the legendary period, and one who seems to know very
little about what she is talking about. She has glaringly elided various
genres not connected to commensality—notably threnody which is
often conceived as a source of comfort—and the Greek cliché that
music did banish angst.32 By focusing on commensal contexts for
song she speaks of occasions from which nurses would have been
socially excluded.33 The nurse’s musicological limitations may be
confirmed by the clear parallels between her words and Medea’s
comments at 298–9:
σκαιοῖσι μὲν γὰρ καινὰ προσφέρων σοφά
δόξεις ἀχρεῖος κοὐ σοφὸς πεφυκέναι.
For if you proffer new pieces of cleverness, to the maladroit your nature
will appear useless and not clever.
Four concepts from the nurse’s words—invention, uselessness, lack
of cleverness (σοφός), and being maladroit (σκαιός)—recur together,
just a hundred lines later. But whereas the nurse decries a set of
musical inventors as useless and maladroit rather than clever,
Medea decries the maladroit who see clever inventors as useless and
lacking cleverness. If Medea is right, the nurse appears both maladroit
(in judging inventions as useless and lacking cleverness) and hypo-
critical (in labelling the inventors as themselves maladroit).34
In summary, by giving the nurse’s words metapoetic force while
undermining her authority, Euripides leaves us in a quandary about
how to interpret his own musical aspirations. The passage will turn
out to prime us both for self-reflexive comments about music, and for
their interpretative complexities.
32
The cliché: Hes. Th. 55 with West (1966); Crane (1990) suggests that Euripides
and his audience might have rejected the cliché, though this is different from the nurse
neglecting it. Actual therapeutic uses of music have a marginal presence in sources
from the late fifth century on: West (2000) 55–66. Pleasure from threnody: e.g. LfgrE
s.v. γόος B4.
33
The nurse mentions ‘songs on many strings’ (196–7), not as a future innovation
which she expects to have therapeutic potential but as a pre-existing state which has
failed to be therapeutic. As Mossman (2011) notes, the usage contrasts with Plato Rep.
1.339cd, where use of ‘many strings’ goes hand in hand with frequent modulation as a
dangerous innovation. I hesitate, on the basis of just two passages, to interpret the
nurse as incorrectly designating traditional music using a term associated with
cutting-edge music.
34
Other readings of the intratext are of course possible. One could be more
cautious of Medea’s rhetorical goals, or acknowledge that the nurse judges past
inventions by their results, whereas Medea’s targets judge their novelty per se.
112 Oliver Thomas

The second key passage is the first half of the first stasimon, where
the chorus sing—under Medea’s gaze—a remarkable response to the
first episode, in which they listened to Medea’s speech about the
humble social position of women, promised to keep silent if an
opportunity for revenge should arise, witnessed the exchange with
Creon, and heard Medea’s plan for violent revenge:
ἄνω ποταμῶν ἱερῶν χωροῦσι παγαί, [str.]
καὶ δίκα καὶ πάντα πάλιν στρέφεται·
ἀνδράσι μὲν δόλιαι βουλαί, θεῶν δ᾽
οὔκετι πίστις ἄραρε,
τὰν δ’ ἐμὰν εὔκλειαν ἔχειν βιοτὰν στρέψουσι φᾶμαι· 415
ἔρχεται τιμὰ γυναικείῳ γένει·
οὐκέτι δυσκέλαδος φάμα γυναῖκας ἕξει. 420
μοῦσαι δὲ παλαιγενέων λήξουσ’ ἀοιδῶν [ant.]
τὰν ἐμὰν ὑμνεῦσαι ἀπιστοσύναν.
οὐ γὰρ ἐν ἁμετέρᾳ γνώμᾳ λύρας
ὤπασε θέσπιν ἀοιδὰν 425
Φοῖβος ἁγήτωρ μελέων· ἐπεὶ ἀντάχησ’ ἂν ὕμνον
ἀρσένων γέννᾳ. μακρὸς δ’ αἰὼν ἔχει
πολλὰ μὲν ἁμετέραν ἀνδρῶν τε μοῖραν εἰπεῖν.
The springs of sacred rivers are running uphill, and justice and the universe
are turning backwards: men’s counsels are deceitful—pledges of the gods
no longer hold fast—and tales will turn my life around to have good repute.
Honour is coming to the womanly race: no longer shall a tale of unpleasant
din constrain women.
The Muses of aged singers will leave off harping on my infidelity. For it was
not within our mind that Phoebus, leader of melodies, bestowed divine
song to the lyre, since I would have sounded a song in response to the
species of males. The long ages have much to say of our lot and of men’s.
The world is topsy-turvy: men are abandoning their oaths; honour
and repute are coming to women, and future music will have to treat
male and female infidelity even-handedly.
Like the nurse, the chorus call for a revolution in music from which
they have been excluded—here a revolution of content and social effect,
rather than psychological effect. The implication that traditional poetry
confers repute and honour upon men puts epic and epinician among
the targets, and the dactylo-epitrite metre supports both connections.35
35
Hopman (2008) 157 notes that Medea featured in early epic Argonautica stories
and in Pythian 4.

The words ὤπασε θέσπιν ἀοιδάν / Φοῖβος (425–6) are a specific allusion
to Od. 8.498 θεὸς [Apollo: cf. 8.488] ὤπασε θέσπιν ἀοιδήν, and indeed
Demodocus’ best-loved song is about the infidelity of Aphrodite.
ἡγήτωρ (426) is mostly confined to epic. Epic did indeed discuss female
infidelity (e.g. of Helen and Clytemnestra; more generally in Hesiod’s
Works and Days); but the contraction and correption of ὑμνεῦσαι (423)
may point also to the dialect of iambus and so of Semonides’ mis-
ogyny.36 Jason accidentally adds a further target during the second
episode, when he cites the ὕμνοι (as in 192, 427) of his fellow-Argonaut
Orpheus as ideal songs whose power would nevertheless not compen-
sate for a loss of honour (542–4). Orpheus’ assassination by barbarian
women was a common fifth-century story, and he turns (by the
Hellenistic period at least) into a misogynistic pederast.37 Euripides
works him in to represent lyre-playing as a male preserve, and to
connect this world with Jason’s honour-code which will come unstuck,
thanks to another barbarian woman.
However, the chorus’s rhetoric cannot straightforwardly be
endorsed. Their radicalism takes linguistic shape when they attribute
separate βιοτή, γένος/γέννα, γνώμη, and μοῖρα (‘life’, ‘race/species’,
‘mind’, ‘lot’) to the two sexes, rather than to the human species as a
whole; ‘I’ repeatedly means a monolithic ‘womankind’. A more con-
servative audience’s suspicions are enhanced by the palpable shallow-
ness in their argument. Since they begin with a proverbial adynaton,
their predictions seem anchored in fantasy. They concentrate on the
‘deceitful counsels’ of men, after a speech emphasizing Medea’s
intention to use deceit—a female speciality (408–9)—against Creon
and Jason. Their inference from the existence of male oath-breaking
(by Jason) to imminent honour for women is unsubstantiated.
Medea’s presence casts doubt on how far they believe their own
words. Further objections arise from outside the characters’ realm
of knowledge. Even if this ode and the first pair of the second
stasimon do discuss male infidelity, the tragedy as a whole will
certainly not bathe women in glory. The claim that lyre-playing is
36
For the linguistic detail see Mastronarde (2002) on 423, 426–7.
37
See in general Gantz (1993) 721–5. Death by Thracian women: Aeschylus’
Bassarids in Ps.-Erat. Cat. 24, LIMC Orpheus section IV. His lyre-playing pitted
against Sirens: A.R. 4.891–917. Pederasty after failing to rescue Eurydice: Phanocles
fr. 1; Ov. Met. 10.78–11.43 (leads Thracian women to kill him).
114 Oliver Thomas

an exclusively male preserve had for the audience at least one famous
counter-example in Sappho.38 And so on.
Confirmation that the chorus deserve scepticism comes in 1085–9,
where they backtrack on their rhetoric in response to Medea’s
anguished monologue. This time Medea is not watching since she
has ominously followed her children indoors; some audience mem-
bers may well expect her to kill them immediately, during the chor-
us’s words.39 They say: ‘I have often approached harder debates than
women should . . . ’
ἀλλὰ γὰρ ἔστιν Μοῦσα καὶ ἡμῖν, 1085
ἣ προσομιλεῖ σοφίας ἕνεκεν,
πάσαισι μὲν οὔ, παῦρον δὲ γένος
(<μίαν> ἐν πολλαῖς εὕροις ἂν ἴσως)
οὐκ ἀπόμουσον τὸ γυναικῶν.
But since we too have a Muse who accompanies us for the sake of our
wisdom40—not all of us, but small is the tribe of women (perhaps you
would find one among many) who are not unfamiliar with the Muses— . . .
. . . I declare: the childless are more fortunate than parents; they
circumvent efforts which may well turn out fruitless.’ The references
to women’s relationship with the Muses and to a female γένος recall
the first stasimon. But though the point here is still radical—young
Greek women rejecting motherhood—the rhetoric has lost the polar-
ization of musical men versus unmusical women, begins with an
apology for earlier bold claims, and is syntactically embarrassed.
Their glaring avoidance of Medea’s situation suggests that their
horror at her plan is a factor in their climbdown.41
So far I have argued that the first stasimon predicts a content-based
revolution in music in terms which are heavily ironized. What if
Clearchus was right to believe that Euripides was innovating through
widespread use of melodic responsion? The first stasimon would be a
crucial moment either to introduce the effect or to confirm that it was
not an isolated strategy of the parodos. Either way, the coincidences
38
On female musicianship in Greece see e.g. Snyder (1989); Greene (2005); Vazaki
(2003) treats fifth-century Athens; Pomeroy (1977) treats Hellenistic developments.
39
In the event, she waits to revel in the messenger’s report of the princess’s death.
But at 1065 she expresses confidence that the death is underway, and gives us no
reason to expect that she will await confirmation.
40
Either ‘because we have wisdom’ or, as Kovacs (1994), ‘to promote wisdom in us’.
41
Cf. Mossman (2011) 332–5.

between form and content would be remarkable. The adynaton
chosen implies a confusion in the natural states of up and down.42
The repetition of πάλιν στρέφεται . . . στρέψουσι suggests the word
‘antistrophe’ itself. ἀντηχέω is both ‘to counter in singing’ but also
‘to sing responding frequencies’.43 And the first line of the antistro-
phe, ‘The Muses of aged singers will leave off . . . ’, is the perfect
sentiment to accompany formal innovation.44
These coincidences of content would still leave us in a similar
quandary about how to interpret Euripides’ musical comments.
The late Ian MacAuslan suggested to me that Euripides, by making
the melody sound ‘wrong’ (cf. 420 δυσκέλαδος?), could reinforce the
chorus’s claim to have been excluded from (lyre-)music. More plaus-
ible, I think, is to attribute the potential topsy-turvy melodization to
the characters, as a means of imitating content in melody, as is found
elsewhere in Greek musical documents.45 Melody would still contrib-
ute to undermining their ideas, in that the audience remains simul-
taneously aware of the disjunction between Euripides’ musical
innovation in form versus the chorus’s prediction of musical innov-
ation in content, which Euripides will not provide.
The final key passage for our purposes is the first strophic pair of
the third stasimon, which occurs just after the chorus have witnessed
Aegeus offering Medea a refuge in Athens, and then heard her plan to
kill her children to exact revenge on Jason.
Ἐρεχθεΐδαι τὸ παλαιὸν ὄλβιοι [str.]
καὶ θεῶν παῖδες μακάρων, ἱερᾶς 825
χώρας ἀπορθήτου τ’ ἄπο, φερβόμενοι
κλεινοτάταν σοφίαν, αἰεὶ διὰ λαμπροτάτου
βαίνοντες ἁβρῶς αἰθέρος, ἔνθα ποθ’ ἁγνὰς 830
ἐννέα Πιερίδας Μούσας λέγουσι
42
Pherecrates’ character Music describes Kinesias’ innovative modulations in
terms of a similar confusion of left and right: fr. 155.11–12.
43
So of sympathetic frequencies in the Aristotelian Problemata (19.24 919b16).
44
For the uses of ‘new’ and ‘old’ in musical discourse of the period see e.g.
LeVen (2014) 87–101. Phonetic correspondences between strophe and antistrophe
are discussed, rather haphazardly, by Irigoin (1988). These may have reinforced the
audience’s perception of melodic responsion.
45
See e.g. in Limenius’ Paean, the double-peak on 2 δικόρυφον, the chromatic run
in 16 αἰόλοις, and the echo of it in 17 πετροκατοίκητος Ἀχώ (numeration follows
DAGM no. 21). Cf. D’Angour (this volume). Cosgrove and Meyer (2006) 74–5
interpret the deliberate opposition of melody and accents in DAGM no. 17 (second-
century CE setting of a threnody for Ajax) as expressing pain.
116 Oliver Thomas
ξανθὰν Ἁρμονίαν φυτεῦσαι·

τοῦ καλλινάου τ’ ἐπὶ Κηφισοῦ ῥοαῖς [ant.] 835
τὰν Κύπριν κλῄζουσιν ἀφυσσαμέναν
χώρας καταπνεῦσαι μετρίας ἀνέμων
ἡδυπνόους αὔρας· αἰεὶ δ’ ἐπιβαλλομέναν 840
χαίταισιν εὐώδη ῥοδέων πλόκον ἀνθέων
τᾷ Σοφίᾳ παρέδρους πέμπειν Ἔρωτας,
παντοίας ἀρετᾶς ξυνεργούς.
Children of Erechtheus, prosperous of old, and offspring of the blessed
gods, you from the sacred and unsacked land who feed on wisdom most
famous, and ever step in luxury through the air most bright—where once,
they say, the nine pure Pierian Muses planted blonde Harmonia;
and they make it known that as the Cyprian drew water at the streams of
fair-flowing Kephisos, she breathed over the land the breezes of winds,
measured and blowing sweetly, and that she ever places a fragrant garland
of rose-flowers on her hair, and sends to sit by Wisdom her Cupids,
collaborators in all forms of virtue.
The chorus presents Athens as largely sublimated from human
needs.46 The Athenians feed not on food but on σοφία—which
includes the skill of the musician. They tread air not soil. Their
stability is stressed not only by ἀπορθήτου (‘unsacked’, 826), but by
the self-responsion of ‘land’ and ‘ever’ (826/837 χώρας, 829/840 αἰεί).
The Athenians are autochthonous children of Erechtheus and gods;
even Aphrodite and the Cupids are desexualized; the Muses ‘plant’
Harmonia without male reproductive involvement.47 The ideology of
autochthony and the tropes of the Golden Age make this an unchan-
ging, pristine Athens.
Clearly the ode has metapoetic potential. Harmonia is a native
Athenian, and Euripides’ song instantiates her; Pucci (1980) 117 adds
that Euripides’ Athenian σοφία is nourishing. However, Euripides is
not simply allowing Athenian choreuts to praise their homeland and
audience. Yet again the characters are given ideas which seem suspect
46
My reading resembles that of Nimis (2007). Swift (2009) 371–5 focuses on the
stasimon’s ironic engagement with locus amoenus tropes. For food, sex, and deteri-
oration as markers of the human condition in Hesiod’s Prometheus myth, see
Vernant (1974) 177–94.
47
The Muses (~ the liberal arts) are logically prior to and generate Harmonia
(musical attunement). The Greek could conceivably mean that Harmonia (social
attunement) planted the Muses, instead of them being born from Zeus’ affair with
Mnemosyne.

from the audience’s perspective and even from their own. The com-
ments on Aphrodite and Sophia seem superficial considering the
ambivalence of these forces in the surrounding play, and the repre-
sentation of Aegeus as both sexually unfortunate and dim-witted.48
The audience would not have advocated stepping ἁβρῶς (‘in luxury’,
829), aerobatically or otherwise, and knew that Athens was no longer
‘unsacked’. Harmonia’s vegetal birth skirts the lust of her normal
parents Ares and Aphrodite.49
Thus once again the characters’ thoughts are undermined both
within the terms of the plot, and by the audience’s broader chrono-
logical perspective. As with the first stasimon, accepting Clearchus’
testimonium would again contribute to the ode’s meaning conveni-
ently. If Euripides was using melodic responsion, the musical form
would, in the very act of performance, deconstruct and expose the
ideological forcedness of the changeable/unchangeable opposition on
which the chorus’s ideas, including their picture of Athenian Harmo-
nia, are premised: Euripides’ Athenian melody innovates by repeat-
ing.50 We have no evidence as to whether Euripides was innovating
here in harmonia considered narrowly (e.g. use of the chromatic
genus, or of modulation), but I do not think it would be hard for an
audience to take Harmonia as also governing the adaptation of her
intervals to accents.
There is, however, a supplementary aspect to Euripides’ potential
melodic innovation and its performative contradiction of the
Corinthian women’s description. Immediately after their idealization,
the chorus wonders whether this city ‘of sacred rivers’ (846 ποταμῶν
ἱερῶν) will receive Medea. This clearly echoes ἄνω ποταμῶν ἱερῶν
from the first stasimon. As I noted above when discussing lines
1085–9, Medea’s plan to kill her children has distanced the chorus
from her. In line with this, they here switch their earlier focus from
the social confusion of Jason’s oath-breaking to that of Medea’s
intended infanticide, and suggest that Medea’s arrival will have an
48
See e.g. Hall (1997) 103; Keen (2009) 628–9 seems too generous.
49
The audience has been put in mind of Demodocus’ song about them at 425 (see
p. 113). Hesiod Th. 933–7 provides a less scandalous version of how they begot
Harmonia.
50
For a deconstructionist approach to the ideology of autochthony see Loraux
(2000), esp. 60, 111–24. Relatedly, Torrance (2013) 224 suggests that Medea’s recur-
rent words for novelty are a metapoetic comment on Euripides’ innovation (if indeed
it was one) of having Medea kill her children.
118 Oliver Thomas

equal potential to turn Athens’ rivers backwards, and to confound the
pristine Athenian Harmonia, in the sense of socio-political attune-
ment, which they have just outlined. This implication allows us to tie
in the final strand of the play’s musical imagery, the widespread
Greek metaphor of civic attunement. At 306–8, Medea continues
her complaints about the lack of respect accorded to inventors by
asking Creon what πλημμελές (literally ‘off-tune’) suffering he fears
from her, then rephrasing the question in terms of injustice. Medea’s
innovative thinking, in Creon’s mind, threatens to introduce a dis-
harmony of crime into the political attunement of Corinth. Later, this
disharmony is revealed to redound on Medea herself too. At 1008 the
tutor describes Medea’s wail as ‘not concordant’ (οὐ ξυνῳδά) with the
news which elicited it, namely that the princess has accepted Medea’s
fatal gifts—the point of no return. At 1269 the chorus predicts that
the gods will send pollution ‘concordant’ (ξυνῳδά) with the crimes of
kin-murderers. The juxtaposition of this claim with another wail—
that of Medea’s children indoors—points, as the tutor had, to the
intrinsic discord of Medea’s plans which harmonize not with her
satisfaction but with the unmusical wailing of her and her children.
The exodos leaves us with Medea accused repeatedly of pollution and
heading off to live with Aegeus at Athens. In this light, lines 846–50
reveal the chorus’s idealization of Athens to be a hyperbole calculated
to oppose stable sophia to Medea’s innovating cleverness, an un-
sacked city to a wrecker of three governing families, a pure Aphrodite
and virtuous Cupids to a passion which leads to pollution, Harmonia
to a force of intrinsic disharmony—in short, to magnify the risk to
Athens which Medea poses.51 The performative contradiction, if the
first strophic pair used innovative melodic responsion, would then
not simply be a matter of showing up the chorus’s naivety, but a
means of making present in music the breakdown which they foresee,
and which they now appreciate may derive from Medea’s pollution
rather than (as in the first stasimon) from men breaking oaths.
I tried first in this essay to clarify some issues in the slippery
testimonium in Athenaeus. I argued that, although certain models
in which Clearchus made a gross mistake are reasonable and cannot
51
Jason implies her pollution at 1346, 1371, 1393, 1406. On the risk of her heading
to Athens (as she reminds us at 1384–5), see Sfyroeras (1994/5). Buchan (2008) 24
sees the chorus’s Athens as a ‘nostalgic fantasy . . . whereas the real city is one that still
reverberates with the consequences of Medea’s arrival’.

be disproved, they are no more plausible than a model in which
Clearchus had justification for thinking that Euripides made a sig-
nificant innovation in tragic melody in the Medea, by employing
melodic as well as metrical responsion, and as a corollary by divorcing
melody from pitch-accents. Secondly, I have argued that musical
discourse is an undervalued but carefully thought-out theme in the
Medea.52 Others have not considered the testimonium in relation to
this material, but it turns out that of the three main passages the first
can be taken as programmatic of Euripides’ handling of the theme,
while the other two are enhanced by the hypothesis of melodic
responsion. The nurse in her parting comment misguidedly longs
for innovation in the realm of music’s psychological effect, and so
primes us for both self-reflexive musical comments and difficulties in
interpreting their degree of dramatic irony. The chorus in the first
stasimon naively predict innovation in the realm of Greek song-
culture’s gender-bias. The hypothesis of melodic responsion fits the
language of the first strophic pair remarkably well, and it could be
interpreted as the Corinthian women choosing inverted melodic
contours in order to express the inversion of the world-order and to
effect the break from musical traditions which they are predicting.
Most complex is the third stasimon, where the chorus project an
idealised, innovationless image onto Athenian Harmonia, before
questioning whether Athens could receive the disharmony of a
child-killing Medea. Again the hypothesis of melodic responsion
interacts intelligibly with the song, showing up the forcedness of the
chorus’s view of Athens, which at first seems simply naïve but is then
revealed as a hyperbole calculated to bring home the risks which
Medea poses to Athenian stability. If only Clearchus had offered
some assertions about how Euripides introduced his innovations—a
few ‘wrong notes’ at a time, or in a sudden revelation (e.g. on the
words ‘The Muses of old singers will leave off . . . ’), or in any number
of possible ways: we might then be able to evaluate the hypothesis
more cogently. As it is, doubtless some readers will still prefer to
reject the testimonium. I hope that I have at least convinced them
that the melodies of the Medea should not only be on musicologists’
wish-lists, since their interaction with the theme of musical innov-
ation probably had far-reaching implications, for example for how
52
I alert readers also to the recent comments on this theme in Gurd (2016) 124–8,
which unfortunately I saw too late to incorporate fully into my analysis.
120 Oliver Thomas

the audience would assess the chorus’ development over the first
three stasima.53
I would like to end by mentioning two modern productions whose
composers have been moved—presumably without Athenaeus’
help—to reinforce the first stasimon’s claims with melodic novelty.
The Japanese director Yukio Ninagawa chose this moment to switch
from recognizably Japanese musical idioms to an orchestral arrange-
ment of Handel’s Sarabande from HWV 437.54 Along with this
musical upheaval, Medea and several chorus-members pull ‘streams’
of red ribbons from their mouths, which Smethurst (2002) 13
explains as in part a reversal of a kabuki trope where faithful girls
place ribbons in the mouths of their boyfriends. Secondly Annie
Castledine’s Greek production in Cambridge in 2007 set Medea
amid a chorus of Edwardian suffragettes. Just before the first antis-
trophe, the composer Elspeth Brooke had the chorus break out of
Greek into English and sing the first verse of the suffragette anthem
‘The March of the Women’.55 These modern theatre-practitioners, if
I have been on the right track, have taken advantage of a performance
opportunity in Euripides’ text which readers had forgotten about
for centuries.
53
I do not have space here to work into my reading the possible interaction of
melodic responsion with another feature of Medea which shows that it was ‘through-
composed’, namely Euripides’ limited metrical palette. All five stasima consist of two
strophic pairs, and in the first four a dactylo-epitrite pair is followed by a predomin-
antly Aeolic one; dactylo-epitrite and Aeolic are also combined in the antistrophic
part of the parodos. The handling of dactylo-epitrite rarely ventures beyond hemiepes
and epitrite elements, while the Aeolic sections return insistently to the colon
x—
˘˘ — ˘ — —, and incorporate iambs and dactyls in recurring ways.
54
My information about Ninagawa’s long-lived production is based on Smethurst
(2002) here p. 12, and on two clips currently available on Youtube (https://www.
youtube.com/watch?v=UQXkmRYag94 and https://www.youtube.com/watch?v=
GTD17KkJ9TQ). Smethurst refers to Corelli’s La Follia, which is a more ornamented
version of the same folk-dance as Handel’s piece.
55
Words by Cicely Hamilton, music by Dame Ethel Smyth, date 1910.
Mesomedes’ Hymn to the Sun

The Precipitation of Logos in the Melos
Stelios Psaroudakes
The Suda informs us that Mesomedes of Crete was active as a

composer in the time of Emperor Hadrian, who reigned from
117–138 AD. Initially a slave, he was freed by the emperor, became a
celebrated poet and composer during Hadrian’s reign, and continued
to be well known in later imperial times. A cenotaph was erected in
Rome in his honour about a century later by Emperor Antoninus
(Caracalla, who reigned 198–217 AD).1 A number of other ancient
authors refer to Mesomedes, while the manuscript tradition has
brought down to us nine of his song lyrics and three (or possibly
four) of his songs.2 This chapter focuses on one of his surviving songs,
the Hymn to the Sun. Other pieces in this volume explore texts whose
music has not survived, or about whose musical setting we have only
the scantiest information.3 As in the case of the ‘Seikilos song’,4 the
survival of the words and melody of Mesomedes’ composition give us
the opportunity to look in detail at their interaction.
1
Suda μ.668 s.v. Μεσομήδης (Adler 3:367.8).
2
For a brief biographical summary see West (1992a) 384, and for a more extended
account Bélis (2003a). For a possible attribution to Mesomedes of the ‘Berlin paean’
(DAGM 166–9, No. 50), see Bélis (2003b) 556.
3
See esp. Phillips, Thomas, D’Angour (this volume).
4
D’Angour (this volume) pp. 64–72.
122 Stelios Psaroudakes

Ὕμνος εἰς Ἥλιον Hymn to the Sun6
Χιονοβλεφάρου πάτερ Ἀοῦς, 1 Father of Dawn with the snowy [a]

ῥοδόεσσαν ὃς ἄντυγα πώλων eyelashes, who, the rosy chariot at
πτανοῖς ὑπ’ ἴχνεσσι διώκεις, the foals’ soaring steps you drive,
χρυσέαισιν ἀγαλλόμενος κόμαις delighting in your golden hair
περὶ νῶτον ἀπείριτον οὐρανοῦ about the limitless back of the sky
ἀκτῖνα πολύστροφον ἀμπλέκων twining your ever-circling beam,
αἴγλας πολυδερκέα πάναν5 the thread of radiance
περὶ γαῖαν ἅπασαν ἑλίσσων, round the whole earth winding,
ποταμοὶ δὲ σέθεν πυρὸς ἀμβρότου while rivers of immortal fire
τίκτουσιν ἐπήρατον ἁμέραν. give birth to lovely day.
Σοὶ μὲν χορὸς εὔδιος ἀστέρων 11 For you chorus serene of stars [b]
κατ’ Ὄλυμπον ἄνακτα χορεύει dances—Lord of Olympos,
ἄνετον μέλος αἰὲν ἀείδων a leisurely song ever singing
Φοιβηίδι τερπόμενος λύρᾳ delighting in Phoibos’ lyre,
γλαυκὰ δὲ πάροιθε Σελάνα 15 and the pale Moon in front [c]
χρόνον ὥριον ἁγεμονεύει leads time and season on
λευκῶν ὑπὸ σύρμασι μόσχων· with her white heifers’ drawing;
γάνυται δέ τέ <σ>οι νόος εὐμενὴς 18 and your benevolent heart is glad, [d]
πολυείμονα κόσμον ἑλίσσων. as it keeps the richly arrayed
universe revolving.
Using highly poetic Doric diction, Mesomedes here paints a magnifi-

cent picture of the Sun god, who with reins in hand leads his brilliant
chariot as he, master of the universe, rises gently to the dome of
the sky;7 while at the same time the moon Selene, tired and pale,
5
Πάναν in DAGM 97 but πάγαν in DAGM 96; πάναν (‘thread’) in West (1992b)
304–5 and West (1992a) 10; παγὰν in Pöehlmann (1970) 16, 17.
6
Translation based on that of West (1992a) 304–6. For an alternative English
translation see Landels (1999) 256.
7
The metope in relief from the Temple of Athena at Ilion in Austin (2005) 208
vividly portrays the scene. For comparable, earlier, iconography on the rising Sun see
the red figure crater attributed to the Leningrad Painter depicting Helios (sun disc
over his head) on his chariot as he travels from East to West over the sea (dolphin) in
Kakrides (1986b) 227 fig. 101; the red figure calyx crater depicting Helios on his
chariot amidst the stars, appearing as boys, plunging into the sea as he rises in id., 197;
and the Apulian crater attributed to the Painter of the Underworld depicting Helios,
Eos, and Phosphoros journeying over the sea in id., 231 fig. 104. For Eos spreading
her saffron robe over the face of all the earth cf. Il. 24.695, Ἠὼς δὲ κροκόπεπλος
ἐκίδνατο πᾶσαν ἐπ’ αἶαν.
Mesomedes’ Hymn to the Sun 123

abandons the heavens on her oxen-driven chariot.8 The stars unite
their voices in song accompanied by the lyre: a hymn to the glory of
His Lordship. The god’s heart is filled with joy and euphoria, struck
by the effervescent cosmic beauty.
What remains to be seen is how the composer dresses his words in
musical sounds (notes, durations, syntax, phrases, motifs, cadences,
etc.) and what use he makes of melorhythmic devices. In the follow-
ing analysis, based on the reading by DAGM, each line of text is
shown with the natural syllabic quantities marked below it;9 above the
verse are the note sign/s ascribed to each syllable (sēmeia), and above
those the durations (chronoi) of the signs10 with their ‘modifications’
(chronōn pathē).11
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ ⌙ –
C C C C I C P C Φ∩ C (∩ proposed by DAGM instead of / in V)
Χι-ο-νο-βλε-φά-ρου πά-τερ Ἀ-οῦς, 1
(σ σ Σ σ σ Σ σ σ Σ Σ)
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ υ – –
Φ Μ Μ Μ Μ Ι C Φ ΜΙ∩ Μ
ῥο-δό-εσ-σαν ὃς ἄν-τυ-γα πώ-λων 2
8
The red figure pyxis lid attributed to the Painter of the Lid in Kakrides (1986b)
225 fig. 100 shows most imaginatively the endless ‘chase’ between Helios and Selene.
For another two other depictions of Selene, see the red figure Boeotian calyx crater
attributed to the Painter of the Paris Judgement of Paris depicting Selene (moon
crescent next to her head) on her chariot travelling over the sea (dolphin) in Kakrides
(1986b) 230 fig. 103 and the red figure kylix attributed to the Brygos Painter depicting
Selene (full moon over her head, two stars on either side) on her chariot in id., 229
fig. 102. In none of the aforementioned Selene depictions are the animals associated
with her oxen, as described in Mesomedes’ hymn, but horses, with or without wings.
Apart from horses and oxen yoked to Selene’s chariot, deer are also mentioned in
mythology (Kakrides (1986b) 230).
9
σ for light/bracheia (pronounced in one unit of time), and Σ for heavy/makra
syllable (pronounced in two units of time when ‘in rhythm’).
10
υ for a single unit (monosēmos); – for a double unit (makra disēmos); ⌙ for a
triple unit/makra trisēmos.
11
Dot (stigmē) for upbeat (arsis), absence of dot for downbeat (thesis). Arsis and
thesis are two of the four ‘modifications’ (pathē) of durations (chronoi), according to
Aristides Quintilianus De mus. 1.13 (see Winnington-Ingram 1963: 31.9–10), the
other two being sound (psophos) and silence (ēremia). The leimma (∩) indicates
prolongation of the note duration by a time-unit. Thus, a long syllable (Σ) sung to
one note, say Φ, will last two units of time (Φ/Σ = –), while the same syllable sung to a
note followed by the leimma will last three units of time (Φ∩/Σ = ⌙); see l. 1, above.
–̇ υυ υ̇ υ̇ – υ̇ υ̇ ⌙ –
Μ ΙΜ Ι Ι Π Μ Ι Ζ∩ Ζ
πτα-νοῖς ὑπ’ ἴχ-νεσ-σι δι-ώ-κεις, 3
(Σ Σ σ σ Σ σ σ Σ Σ)
–̇ υυ υ̇ υ̇ – υ̇ υ̇ – υ –
Μ ΖΜ Ζ Ι Μ <Ι> Ι Μ Ζ Ι (<Ι> added by DAGM)
χρυ-σέα͜ ι-σιν ἀ-γαλ-λό-με-νος κό-μαις 4
(Σ Σ σ σ Σ σ σ Σ σ Σ)
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ – υ υυ
Μ Ι Ζ Ι Μ Ι Π Φ C Ρ ΡC
πε-ρὶ νῶ-τον ἀ-πεί-ρι-τον οὐ-ρα-νοῦ 5
(σ σ Σ σ σ Σ σ σ Σ σ Σ)
υ̇υ ̇ – υ̇ υ̇ – υ̇ υ̇ – υ –
CΡ Μ C Μ Μ Μ Μ Μ Ι Μ
ἀκ-τῖ-να πο-λύσ-τρο-φον ἀμ-πλέ-κων, 6
–̇ – υ̇ υ̇ – υ̇ υ̇ υ– υ υ
Ι ̄ Μ Ρ Μ Ι Ζ Ι ΜΡ∩ Ρ C (∩ proposed by DAGM instead of / in N)
αἴ-γλας πο-λυ-δερ-κέ-α πά-ναν 7
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ ⌙ –
C Ρ Μ Μ Μ C R Φ Μ∩ Μ (∩ added by DAGM)
πε-ρὶ γαῖ-αν ἅ-πα-σαν ἑ-λίσ-σων, 8
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ – υ –
Μ Ι Ζ Ζ Ζ Ζ Ζ Ε Ι Ε Ζ
πο-τα-μοὶ δὲ σέ-θεν πυ-ρὸς ἀμ-βρό-του 9
–̇ – υ̇ υ̇ – υ̇ υ̇ – υ –
Ρ Μ Ι Ζ Ζ Ι Μ Ρ <Ι> C (<Ι> added by DAGM)
τίκ-του-σιν ἐ-πή-ρα-τον ἁ-μέ-ραν. 10
–̇ – υ̇ υ̇ – υ̇ υ̇ – υ –
C Φ C P M MM P P C
Σοὶ μὲν χο-ρὸς εὔ-δι-ος ἀσ-τέ-ρων 11
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ ⌙ –
Μ Ι Μ Μ Ι Ρ Μ Ι Ζ∩ Ζ (∩ added by DAGM)
κατ’ Ὄ-λυμ-πον ἄ-νακ-τα χο-ρεύ-ει 12
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ ⌙ –
Z Z M Z Z M Z I E∩ Z (∩ added by DAGM)
ἄ-νε-τον μέ-λος αἰ-ὲν ἀ-εί-δων 13
–̇ – υ̇ υ̇ – υ̇ υ̇ – υ –
Μ Ι ΖΖ Μ Ι Ρ Φ Ζ Ζ
Φοι-βη-ί-δι τερ-πό-με-νος λύ-ρᾳ. 14
υ̇υ ̇ – υ̇ υ̇ – υ̇ υ̇ υ– –
CP M M M C P M MI∩ M
Γλαυ-κὰ δὲ πά-ροι-θε Σε-λά-να 15
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ ⌙ –
Ι Μ Ι Μ Μ Ρ Μ Ι Ζ∩ Ζ (∩ added by DAGM)
χρό-νον ὥ-ρι-ον ἁ-γε-μο-νεύ-ει 16
υ̇υ ̇ – υ̇ υ̇ – υ̇ υ̇ υυυ –
ΜΙ Ζ Ι Μ Ι Φ C ΡΜΡ C
λευ-κῶν ὑ-πὸ σύρ-μα-σι μόσ-χων· 17
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ υυ υ –
C C C C C C P C PΦ Ρ Μ
γά-νυ-ται δέ τέ σοι νό-ος εὐ-με-νὴς 18
υ̇ υ̇ – υ̇ υ̇ – υ̇ υ̇ υυυ –
Μ Ι Ζ Ι Μ Ι Φ C PMP C
πο-λυ-εί-μο-να κόσ-μον ἑ-λίσ-σων. 19
Metrically speaking, there is a sequence of two types of colon con-
sisting of paroemiacs and apocrota.12 These cola have the same form
to begin with but differ in their endings. Both comprise lyric ana-
paests,13 apart from the pentaseme ending (σΣ) of the apocroton:14
σσΣ σσΣ σσΣ Σ –– –– –––– ––––||
paroemiac
˘˘
––
˘˘ ˘˘
σσΣ σσΣ σσΣ σΣ apocroton
˘˘––˘˘––˘˘––˘––|| || = period end
12
The latter was a common metrical form in the second and third centuries AD:
West (1992b) 305 n. 8.
13
Pöehlmann and Speliopoulou (2007) 139.
14
West (1992a) 305 with n. 8; Pöehlmann and Speliopoulou (2007) 139–40.

The first question to be tackled is how the metrical quantities of the
syllables are treated in the melodic rhythm. Do they maintain the
temporal values they have in speech, or are at least some of them
changed by being either prolonged or shortened? An answer may be
given by examination of the rhythmical notation. The observed
rhythmical features involve the allocation of:
1. a monoseme note to a short syllable;15
2. a diseme note to a long syllable;16
3. two monoseme notes to a long syllable.17
The most interesting rhythmic feature, however, is the allocation of a
triseme temporal value to the first of the two long syllables of the
ending of the paroemiacs. This is done in three different ways, either
by (1) allocating one triseme note to it,18 or (2) one monoseme note
followed by a diseme note,19 or (3) three monoseme notes.20 In this
way, the endings of the eleven paroemiacs last as long as the endings
of the eight apocrota, both endings being of a pentaseme duration. In
the second case, that of the two notes (υ–/Σ), the diseme length is
notated with the use of the leimma (∩), a rhythmic symbol which
indicates the protraction by a monoseme duration of the note to
which it is appended.21 This observation raises a further question:
why mix two different cola, one with a tetraseme and another with a
pentaseme (metrical) ending, if the intention is to equate their
rhythms after all, especially if the use of a second type of colon will
not at the end offer any rhythmic variety to the song? In other words,
the prolongations of the long syllables at the ends of the paroemiacs
are not made for any aesthetic effect, since they all occur at the
same place in the cola; they seem only to offer a regularization of
the rhythm.
We are fortunate to possess practically the whole of the song text
(verbal and musical), and thus have before us the complete scale of
the piece. The melodic symbols are found as a group in the Lydian
15
υ/σ, e.g. l. 1: P/πά (in πάτερ). 16
–/Σ, e.g. l. 1: C/ρου (in χιονοβλεφάρου).
17
υυ/Σ, e.g. l. 3: ΙΜ/νοῖς (in πτανοῖς). 18
⌙/Σ, e.g. l. 3: Ζ∩/ώ (in διώκεις).
19
υ–/Σ, e.g. l. 7: ΜΡ∩/πά (in πάναν).
20
υυυ/Σ, e.g. l. 17: ΡΜΡ/μόσ (in μόσχων).
21
What we do not encounter in this song are: the allocation of a diseme note to a
short syllable (–/σ), or of a monoseme note to a long syllable (υ/Σ)—in any case, there is
no way this could be notated –, or of a diseme note followed by a monoseme note (–υ/Σ).

tropos, its diatonic variety.22 In a rising sequence they are:
RΦCΡΜΙΖΕ, an eight-note scale spanning an octave, from the par-
hypatē hypatōn (R) to the tritē diezdeugmenōn (Ε).23 The song begins
on hypatē mesōn (C), a note which occurs most frequently in the
melody, and which also happens to be the very last note of the piece,
creating to our ears the impression that the note functions as a tonal
centre.24 Every line/colon, is treated as a melodic phrase, which
usually ends at the hypatē hypatōn C, or the diatonos mesōn
M. A few phrases end on paramesē Z.25
In section α.1–5 (Χιονοβλεφάρου–οὐρανοῦ), there is a gradual rise of
the tonal level of the melody, from phrase to phrase,26 from the
beginning of the piece to the end of l. 5 (οὐρανοῦ), which, one might
say, paints in sound the progressive rise of Helios’ chariot across the
heavens (Figure 5.1). Although this part (the first five lines) comprises
a melodic unit, the same is not true for the corresponding semantic
unit, which continues to the next line (l. 6) and is integrated at the end
of it (ἀμπλέκων). In other words, melodic cadence (end of l. 5) and
semantic close (end of l. 6), in this first part of the hymn, do not
coincide. Perhaps the exact correspondence of the two phrases
(melodic and semantic) is sacrificed to the description in sound of
the movement of the Sun’s chariot: the melody must reach its peak at
the word οὐρανοῦ, a word standing for the apex of the heavenly dome.
In part α.6–10 (ἀκτῖνα–ἁμέραν) a second melodic cadence (end of
l. 7 πάναν), also does not find a corresponding semantic close, the
latter being delayed by a whole line (the end of l. 8, ἑλίσσων).
However, at the next firm melodic cadence on the tonal centre
C (end of l. 10 ἁμέραν) the meaning is also completed. The melodic
phrase which begins on l. 8 (περὶ γαῖαν) rises tonally and ‘swells’ like a
wave in l. 9 (ποταμοί) upon an insistently repeated Ζ, before it gently
22
For the musical signs (sēmeia) of the fifteen tropoi systēmatikoi of the
Alypian notation system, see MSG 367–406, and for the Lydios diatonikos tropos,
ibid. 368–9.
23
For the names of the notes (dynameis) in the tropos see West (1992b) 222 fig. 8.1.
24
So Winnington-Ingram (1936) 42; Hagel (2009) 289.
25
Winnington-Ingram (1936) 43 comments: ‘Although the principal notes are
standing notes, there is little emphasis on the tetrachord. Indeed to the modern ear it
seems that a feeling for the triad is clearly shown; comp. Nemesis and Seikilos.’
26
Numbers at the top of Figure 5.1 signify the beginnings of lines/cola.

and sweetly makes an inescapable cadence on the tonal centre
C (ἁμέραν).27 Thus the gestation of the day has come about; the rivers
of fire have at last given birth to light, and the melodic tension caused
by the insistence of Z has resolved to peace and quiet.28
In the next section (β, ll. 11–14, Σοὶ–λύρᾳ), a second melodic unit
begins with l. 11. The climate here is changed: it becomes joyful and
encomiastic of the god. The melody begins low, at hypatē hypatōn
(C), and soon rises to the level of paramesē Z (l. 12), the note
appearing insistently during the next two lines (ll. 13, 14). These
lines give a sense of the lively movement and excitement generated
by vivid song and dance, representing to aesthetic effect the words
χορός, χορεύει, ἀείδων, τερπόμενος λύρᾳ. The melody manifests an
agitation and a swirling typical of joyful dance. The melodic phrase
unexpectedly, with a high leap up from below (ΦΖ), hits paramesē (Z)
again, a note insistently visited in this section—leaving an image, it
might seem, of the dancer’s foot raised up and hanging in still air.
A third melodic unit (γ) seems to begin with l. 15 (Γλαυκὰ), and
finish, together with the corresponding verbal meaning, at the end of
l. 17 (μόσχων). In this part, Selene is described seated on her chariot,
which is drawn along the sky by white oxen. Three melodic phrases
make up this section, closely related to each other tonally, giving
the impression of a unit. The melody emanates quietness, calm, and
tiredness, states which are appropriate to the recovering Selene, but
also to the feeling of night.29 Here for the first time the last but one
syllable of the ending scheme ΣΣ is given to three (as opposed to one
or two) notes. There follows a deep melodic cadence which echoes the
sinking of the Moon. Why is it only now, in the penultimate line of
the piece, that the composer chooses to employ three monoseme
notes, one may wonder? An attempt to answer this question will
follow shortly.
27
This feeling is generated if the two successive falling intervals are rendered in
performance in a gentle and sweet vocal character. Winnington-Ingram (1936) 42
characterizes the break at ἁμέραν as ‘important’.
28
The ‘clearly dissonant tritone’ (Hagel (2009) 237) ZP at the junction of lines 9
and 10 contributes significantly to the generation of these emotions.
29
The rendering of the first α of Σελάνα ας υ–/α, creates a ‘yawning’ effect, if the
voice is gently slurred during performance, as it ascends from the first note to its close
neighbour. Also, the falling melody which concludes the section, performed in a
hesychastic manner, will vividly convey the ‘sleepy’ atmosphere.

The last melodic unit (δ, ll. 18–19, Γάνυται–ἑλίσσων) constitutes
an expression of courtesy towards the sun god (Helios), who experi-
ences delight as he rotates the fair world through space. The two
melodic phrases of the unit are in close tonal relationship with each
other, with the second seeming to answer to the first. They both
move and develop at a low level, despite the momentary upward
‘wave’ of the second phrase (beginning of l. 19), leading the song
back to where it started from, the hypatē hypatōn C, the tonal centre.
Mesomedes chooses to rhythmicize ἑλίσσων, a word which expresses
a whirling motion, in three monoseme notes. This is what he does
just two lines previously with μόσχων (17). Undoubtedly, the use of
the triple motif suits well the notion of the rotating world, but it also
functions as a closural signal—this motif has not been used thus
far—a winding up of the whole melos, drawing the listener’s atten-
tion to the fact that the song is about to end.30 In that case, it may
very well be that the three-note motif in the earlier phrase (l. 17)
prefigures the ending, because the use of the motif there is not
justified aesthetically, not being connected to the meaning of the
word μόσχων. Thus, the triple motif seems to have been intention-
ally kept by the poet for the end, a final gesture from his stock of
aesthetic surprises.
Some concluding remarks may be offered about the hymn’s styl-
istic traits. Short syllables are always set to a short note (υ/σ), and
long syllables either to a diseme note (–/Σ), or to two monoseme
notes (υυ/Σ). It is only the penultimate (long) syllables of the
paroemiacs which are extended to a triseme value, receiving either
a single triseme note (⌙/Σ), or a monoseme and a diseme (υ–/Σ)—
never in the reverse order, –υ—or three monoseme notes (υυυ/Σ).
Only some of the signs of the (Lydian diatonic) tropos are employed,
and only of the meizon systema at that; no use of the syneimenon
tetrachord is made. A predominant tonic centre seems to be the
hypatē mesōn (C), for not only does the piece start and finish on this
note, but all the major melodic cadences occur on it (at ends of ll. 5,
8, 10, 11, 17). Secondary melodic endings occur on the diatonos
mesōn (M) and the paramesē (Z). In general, melodic phrases and
colon endings coicide.
30
An ‘end signal’, in ethnomusicological terms.

There is a strong tendency of imitation in the music of the verbal
content (‘word painting’): gentle and steady rise of the melodic line
in response to Helios’ rising chariot (ll. 1–5); melodic tension and
release in consonance with the ‘gestation and birth’ of day (ll. 9–10); a
joyful, undulating melodic line for the chorus of the stars, as they sing
and dance in praise of the god (ll. 11–14); a sleepy, ‘yawning’ (υυ––/
Σελάνα) melody as Selene retreats to her chambers (ll. 15–17); a
triumphant, ‘trumpeting’, monotone melodic ending (υυ–υυ–/γάνυται
δέ τέ σοι) in accord with the joyful mood of the god, as he proudly
confronts the universe (υυυυ–/ἑλίσσων), flooded with his light (ll.
18–19). On two occasions, a strong melodic cadence (on the dominant
tonic centre C) precedes the corresponding verbal phrase by a whole
colon (melody end at 5 vs phrase end at 6; melody end at 7 vs phrase
end at 8). In this phrase, diction seems to be sacrificed in the service of
the music. The mimetic aspects of melos in this composition are
broadly similar to those at work in the ‘Seikilos song’ and, more
distantly, in the chorus of Euripides’ Orestes. Seeing these three texts
alongside each other offers a tantalizing glimpse of continuities in how
ancient composers in widely different eras drew on melodic resources
to produce similar, or at least comparable, mimetic effects.
The musical meaning of this hymn is perhaps revealed only by a
musically expressive rendering in performance.31 It may be therefore
that the present hymn was composed as a recital piece, rather than a
ceremonial solo by a cantor or a priest (as Barker has suggested), in
which case the musical setting would have been less noticeable or
significant.32
31
For a brief discussion of the ethnomusicological terms ‘composition’ and
‘performance’ style, see Psaroudakes (2010) 60.
32
Barker (2002) 148. Along the same lines, West (1992a) 384 finds the Hymn to
the Sun a rather ‘limited and uninspired’ piece of music, and observes ‘no perceptible
correlation of melody to meaning’. Landels (1999) 205 also thinks that Mesomedes
‘would have been a quite insignificant figure but for the strange quirk of fate by which
some of his music has survived in manuscript, copied out many times at various
dates’. Mathiesen (1999) 57 decribes Mesomedes’ hymns as ‘extremely simple, exhib-
iting no modulations . . . and very little rhythmic variety and melodic distinction’.
Hagel (2009) 87 n. 98 calls Mesomedes’ hymns ‘simple music’, no doubt on similar
grounds, that is, lack of formal complexity. Anderson (1980) 196 is even more
derogatory: ‘it is all the more to be regretted that neither his poetry nor his melody
suggests any ability to rise beyond mediocrity’. However, as regards at least the poetic
diction of Mesomedes, Easterling and Knox (1999) 853 attribute to it ‘grace and
subtlety: his art is a product of a sensitive talent’.
Figure 5.1. Melodic diagram of Mesomedes’ Hymn to the Sun.

Figure 5.1. Continued.




Part II
Theory, Reception, Contexts

Hearing the Syrinx in Euripidean Tragedy

Naomi Weiss
In 1927, nine years after Claude Debussy’s death, the editor of his
previously unpublished solo flute piece, La Flûte de Pan, changed its
title to Syrinx. The music was originally composed as incidental
music for Gabriel Mourey’s dramatic poem Psyché, and seems to
belong in Act Three, following the stage direction ‘Sometimes they
[the nymphs] stop [dancing] completely, amazed, listening to the
syrinx of invisible Pan, moved by the song that escapes from the
hollow reeds’ (‘Par moments elles s’arrêtent toutes, émerveillées,
écoutant la syrinx de Pan invisible, émues par le chant qui s’échappe
des roseaux creux’).1 The music they hear is actually that of the flute
representing the sound of the god’s panpipes, his syrinx, which is also
sometimes referred to as a flute in Mourey’s text. The change of the
piece’s title is commonly thought to refer to a myth that had little to
do with Mourey’s play—the story of the nymph Syrinx, who, running
from Pan’s amorous advances, is transformed into reeds, from which
he fashions his instrument.2 But the title also encourages us still to
hear the flute as the god’s syrinx, so that, even without the dramatic
1
Mourey, Psyché III.1. On the dramatic context of Debussy’s Syrinx, see Fulcher
(2001) 132–3. It is clear from Debussy’s correspondence with Mourey that the piece
was intended to be heard at a particular moment in the play: in a letter dated
17 November 1913 he writes, ‘So far I have not found what is needed . . . since a
flute singing on the horizon must at once contain its emotion! . . . Tell me, please, very
exactly, the lines after which the music comes in’ (‘Jusqu’à ce jour je n’ai pas encore
trouvé ce qu’il faut . . . pour la raison, qu’un flute chantant sur l’horizon doit contenir
tout de suite son emotion! . . . Dites moi, je vous prie, très exactement, les vers après
lesquels la musique intervient’).
2
As told in Ov. Met. 1.689–712; Longus 2.34; Ach. Tat. 8.6.
140 Naomi Weiss

context, the music can create a mimetic effect, a conflation of two
different but conceptually related instruments.3
A similar sort of instrumental mimesis must have regularly
occurred in the classical Athenian theatre, where, though the aulos
(not a flute, but a double reed pipe) usually provided the musical
accompaniment to the songs of both actors and choruses, other
instruments could be evoked as well. This phenomenon may have
been particularly common in performances of the dithyramb and
satyr play: in Pindar fr. 70b, one of the few surviving examples of
the dithyramb, there is a very vivid acoustic and visual image of the
‘whirlings’ of drums and noise of castanets (ῥόμβοι τυπάνων, / ἐν δὲ
κέχλαδ[εν] κρόταλ’ . . . , fr. 70b S-M 9–10), yet the genre by this point
was ‘decidedly aulodic’;4 in Sophocles’ Inachus, which seems to have
included a scene in which Hermes lulls the many-eyed giant Argus to
sleep by playing on his syrinx, the music to which the chorus react
could have been represented by that of the aulos, the standard
instrumental accompaniment.5 This is not to say that other instru-
ments would never have appeared in the Theatre of Dionysus at
Athens: the fragments of Sophocles’ Ichneutai, for example, include
a long passage in which the satyr chorus fearfully responds to the
sound of the chelys lyre, which could have been represented by the
aulos, but may instead have been produced by a loud, concert kithara
lyre offstage instead.6 This might have occasionally happened in
tragedy as well, especially in plays in which a particular role required
the playing of a lyre (or at least the illusion of it): Sophocles, playing
the title role in the first performance of his Thamyras, apparently
‘took up and played the kithara’;7 a kithara might also have been used
3
Debussy creates the impression of panpipes in other compositions as well,
especially those which concern faun figures, such as the first piano duet in Six
Epigraphes Antiques, of which the theme is ‘Pour invoquer Pan dieu du vent d’été’
(‘To invoke Pan, god of the summer wind’): see Raad (2005) 40.
4
Power (2013) 240. Franklin (2013) argues that the early dithyrambs established
by Arion in Corinth might have been performed to the accompaniment of the kithara:
cf. Koller (1962).
5
On the syrinx as represented by an aulos here, see Power (2012a) 297–8; Griffith
(2013) 271.
6
Soph. Ich. TGrF F 314.124–337, On the possibility of an actual lyre appearing in
some form here, see Power forthcoming. On the intensely musical focus of this
passage, see Griffith (2013) 269–71.
7
Vit. Soph. 5. On the use of a kithara in Soph. Tham., see Wilson (2009) 75; Power
(2012a) 298–30; 2013: 239.
Hearing the Syrinx in Euripidean Tragedy 141

in Euripides’ Antiope, in which Amphion seems to have entered
singing to his lyre (TGrF F 182a), before telling the chorus about
the instrument’s history (TGrF F 190–2). But these two examples
seem to have been exceptions, and in extant tragedy, instead of a
particular instrument being central to a protagonist’s character, dif-
ferent types of musical sound are usually referred to in the choral
songs, which would have been accompanied by the aulos.8 Like Pan’s
syrinx in Mourey’s play, such descriptions of instrumental music,
which are especially common in the later work of Euripides, would be
enacted through the performance of aulete in the theatre. The result
would therefore be a two-way affective interaction between words and
music: the verbal description transforms the instrumental sound heard
by the audience; at the same time the physical nature of that sound (its
tune, tuning, volume, pitch, rhythm, and timbre) can suggest a layer of
meaning not otherwise present in the words themselves.9
Such mimetic effects would have been particularly successful when
the instrument described had some acoustic and/or cultural affinity
with the one performed, as in the case of the urban aulos and the
syrinx, its rustic cousin.10 We can see the close association of these
two instruments in the use of the word syrinx to refer—at least in
texts from the mid-fourth century BC onwards—to a device that could
somehow raise the pitch of the aulos (whether it was some sort of
mechanism fitted to it or like the ‘speaker hole’ of modern woodwind
instruments is disputed).11 Aristoxenus, for example, writes that
‘when the syrinx is pulled down, the highest note of him who plays
the syrinx, compared with the lowest of him who plays the aulos,
would exceed the stated limit’ (τῆς σύριγγος ὁ τοῦ συρίττοντος
ὀξύτατος πρὸς τὸν τοῦ αὐλοῦντος βαρύτατον μεῖζον ἂν ποιήσειε τοῦ
ῥηθέντος διαστήματος, 1.20–1).12 We do not know if an aulete would
8
On the aulos as the primary musical accompaniment of tragedy, see esp. Wilson
(1999) 76, with full bibiography; also Wilson (2008) 185–6.
9
Cf. Phillips in the introduction to this volume on an ‘intermedial’ approach to
ancient Greek music.
10
Cf. Allan (2008) 324–5: ‘If the αὐλός-player in the theatre ever attempted to
imitate other instruments mentioned in a play, the σῦριγξ will have been . . . among
the easiest.’
11
On the nature of the syrinx device, see esp. West (1992b) 86, 102–3; Hagel (2012).
12
Cf. Ps.-Aris. De aud. 804a14; Ps.-Plut. De mus. 1137f4–1138a6; Plut. Non posse
vivi 1096b. συριγμός also refers to the hissing of a snake, and at the Pythian auletic
contest it seems to have traditionally represented the dying serpent killed by Apollo,
for which auletes tended to use the pitch-raising device called a syrinx: see Xen. Symp.
142 Naomi Weiss

have made use of such a device in a tragic performance, but the
vocabulary associated with it (σύριγξ, συριγμός, συρίζω) suggests
that musicians imitated—or, more importantly, were perceived as
imitating—the sound of the syrinx on the aulos.13
The syrinx as an instrument referred to in a play’s text seems to be
a latecomer to Athenian tragedy, mostly appearing in a cluster of
plays by Euripides from the last two decades of the fifth century. The
rarity of references to this long-used instrument in earlier tragedy
may initially appear rather surprising, yet, unlike the aulos and lyre,
the syrinx does not seem to have played a significant part in Athens: it
is rarely represented in vase painting except in the hands of Pan,
suggesting that, for many Athenians, it was associated more with the
rustic imaginary than with everyday urban life.14 The herdsman’s
syrinx must still have been a recognizable sound for the audience of
tragedy, especially for those who came to the theatre of Dionysus
from more rural parts of Attica and beyond. In the archaic period at
least, its music seems to have been part of wedding celebrations as well:
the Muse Calliope plays it on both the François Vase (c. 570 BC) and the
Sophilos Dinos (c. 580 BC).15 But given the general lack of evidence for
this instrument having much prominence in classical Athens, it is
perhaps more surprising that it should appear in any marked way in
tragedy at all. The sudden frequency with which it is referred to in
Euripides’ later plays corresponds, however, with his experimentation
with the much-discussed ‘New Musical’ trends toward the end of the
fifth century, and coincides with an increase in extended descriptions
of μουσική (music, song, dance) in his work in general.16
6.5; Strab. 9.3.10. The term’s technical meaning is therefore inextricably tied to its
acoustic one, which links the sound of the syrinx as an instrument to whistling or
hissing noises.
13
Writing many centuries later, Achilles Tatius compares the two instruments,
seeing the syrinx as a combination of auloi: ‘The syrinx is in reality many auloi, and
each reed is an aulos, while all the reeds together pipe (aulousi) just as one aulos’
(ἡ σύριγξ αὐλοὶ μέν εἰσι πολλοί, κάλαμος δὲ τῶν αὐλῶν ἕκαστος· αὐλοῦσι δὲ οἱ κάλαμοι
πάντες ὥσπερ αὐλὸς εἷς, 8.6.3).
14
On the syrinx in Greek (especially Athenian) life, see West (1992b) 110–12.
15
It also appears in this context in the hymeneal third stasimon of Euripides’
Iphigenia in Aulis (see pp. 157–8), but its inclusion there may be based on the epic
account of the wedding of Peleus and Thetis in the Cypria: see West (1992b) 110 n. 12.
16
On the ‘New Music’ in general, see esp. Csapo (2004); (2011); D’Angour
(2006a); (2007); (2011): 184–206; LeVen (2013); (2014). On Euripides and the ‘New
Music,’ see esp. Csapo (1999–2000); (2008); (2009); Steiner (2011).

The syrinx is mentioned in two extant tragedies prior to the 420s:
Prometheus Bound, of which the authorship and date are far from
certain (though the play seems likely to have been composed before
the final quarter of the fifth century),17 and Euripides’ Alcestis, an
unusual ‘prosatyric’ play to which I will return towards the end of this
discussion. In the former play, Io, in her frantic monody, imagines
Argus is still pursuing her, and hears the sound of pipe-playing:
ὑπὸ δὲ κηρόπλαστος ὀτοβεῖ δόναξ
ἀχέτας ὑπνοδόταν νόμον·
ἰὼ ἰὼ πόποῖ, ποῖ μ’ ἄγουσι τηλέπλαγκτοι πλάναι;
And in accompaniment the clear-sounding, wax-moulded reed booms
forth a tune that brings sleep. Iō, iō, popoi! Where are my far-roaming
wanderings taking me? ([Aesch.] PV 574–6)
These lines must refer to Hermes, who in some accounts is said to
have lulled Argus to sleep by playing his syrinx before killing him.18
As I mentioned above, this performance presumably would have
been staged in Sophocles’ satyr play Inachus: two of the surviving
fragments draw attention to the sound of Hermes’ pipe-playing;19
Lloyd-Jones suggests that Hermes challenged Argus, who apparently
came on stage ‘singing’, to a musical competition.20 In the satyr
play the sound of the aulos could have represented that of Hermes’
syrinx, which the actor might then only mime playing.21 Similarly, in
Prometheus Bound, though it is possible that the syrinx might have
been briefly used as a small visual prop, it is far more likely, given the
brevity of the reference to it here, that Io’s monody would have been
accompanied by the aulos, the main instrument of the theatre. While
hearing this vivid description of the loud pipes with their dangerously
soporific noise, the audience would therefore simultaneously hear the
sound of the aulos, which would momentarily represent the syrinx.
This metamusical moment suggests that the accompanying tune of
17
The terminus ante quem of 424 BC is based on echoes of Prometheus Bound in
Aristophanes’ Knights: see Griffith (1977), esp. 9–13 for a discussion of the play’s date;
cf. Sutton (1983); Flintoff (1986).
18
Cf. Bacch. 19.35–6; Ovid, Met. 1.682–714.
19
Soph. TGrF F 269c7 (σύριγγος δὲ κλύω), 296c21–4 (the chorus guesses that
Hermes is the source of the sound).
20
Lloyd-Jones (1996) 115–16.
21
A suggestion made by Power (2012a) 297–8.
144 Naomi Weiss

the aulos (here imagined as a syrinx) could have played an important
role in Io’s monody, appearing to stir her to madness, or at least
reflecting her frantic state. As Peter Wilson has demonstrated, its tune
assumes a similarly destructive and maddening force in Euripides’
Heracles, from the moment when Lyssa proclaims that she will ‘pipe
you down with fear’ (καταυλήσω φόβωι, 871).22
In a performance, then, the μουσική that Io describes would over-
lap or merge with the μουσική produced by the aulete in the orchestra,
so that, in this combination of syrinx and aulos, the audience would
in effect hear two registers of instrumental sound. We are often
prompted in more modern performances to imagine the sound of
an instrument as something else, whether it be the flute as panpipes,
as in Debussy’s Syrinx, a keyboard glockenspiel as Papageno’s magic
bells in Mozart’s The Magic Flute, or even an oboe as a duck in
Prokofiev’s Peter and the Wolf. Bruce Smith discusses a similar
auditory experience in early modern theatre, in which ‘sound . . . is
important not so much for what it is as for what it signifies. What
audiences actually heard in the theatre and what they imagined they
heard may not always have been the same thing.’23 He gives an
example from Act Five of Shakespeare’s Coriolanus, when the printed
text includes the stage direction for trumpets, hautboys, and drums,
while the messenger describes a much wider range of instrumental and
vocal sound, thereby inviting the audience to hear imagined as well as
physical and perceptual phenomena through a process of metonymy.24
Sound Studies scholar Don Ihde similarly describes an audience’s
reception of a sound in the theatre as a synthesis of imagined and
perceived sound, a form of ‘auditory polyphony’.25 So in Io’s monody
in Prometheus Bound the audience would hear not only the aulos as the
syrinx, but also a combination of both instruments at once.
It is primarily in the later plays of Euripides, particularly the choral
odes, that we find multiple occurrences of such mixing or layering of
the aulos in the theatre and the syrinx in the words of a song. After
Alcestis and Prometheus Bound, the syrinx appears in Electra, which
(if we accept its dating on metrical grounds to around 420) is his
earliest extant tragedy to display a strikingly self-conscious and
extended engagement with μουσική.26 As Eric Csapo has shown, the
22 23
Wilson (1999–2000). Smith (1999) 242 (emphasis original).
24 25
Coriolanus F1623: 5.4.49–52. Ihde (2003) 62.
26
On this dating of Electra, see Cropp and Fick (1985) 23, 60–1; Cropp (1988) l–li.

first stasimon (432–86) includes imagery typical of the ‘New Music’
and new dithyramb, as the chorus sings of the ‘aulos-loving’
(φίλαυλος, 435) dolphin ‘whirling’ (εἱλισσόμενος, 437) around the
prows of the Greeks’ ships as they go to Troy.27 When it describes
Achilles’ armour in the second half of the ode, its language is full of
allusions to choreography, so that the ecphrasis appears to come alive
through its performance. Musical imagery continues in the second
stasimon, which the chorus begins by picturing Pan piping on his
syrinx as he brings the golden fleece from the mountains to Argos
(Eur. El. 699–706):
ἀταλᾶς ὑπὸ {ματέρος Ἀργείων{ [1st Str.]
ὀρέων ποτὲ κληδὼν 700
ἐν πολιαῖσι μένει φήμαις
εὐαρμόστοις ἐν καλάμοις
Πᾶνα μοῦσαν ἡδύθροον
πνέοντ’, ἀγρῶν ταμίαν,
χρυσέαν ἄρνα καλλίπλοκον 705
πορεῦσαι. πετρίνοις δ’ ἐπι-
στὰς κᾶρυξ ἰαχεῖ βάθροις·
Ἀγορὰν ἀγοράν, Μυκη-
ναῖοι, στείχετε μακαρίων
ὀψόμενοι τυράννων 710
φάσματα {δείματα.
χοροὶ δ’{ Ἀτρειδῶν ἐγέραιρον οἴκους.
θυμέλαι δ’ ἐπίτναντο χρυσήλατοι, [1st Ant.]

σελαγεῖτο δ’ ἀν’ ἄστυ
πῦρ ἐπιβώμιον Ἀργείων· 715
λωτὸς δὲ φθόγγον κελάδει
κάλλιστον, Μουσᾶν θεράπων,
μολπαὶ δ’ ηὔξοντ’ ἐραταὶ
χρυσέας ἀρνὸς {ἐπίλογοι{
Θυέστου· κρυφίαις γὰρ εὐ- 720
ναῖς πείσας ἄλοχον φίλαν
Ἀτρέως, τέρας ἐκκομί-
ζει πρὸς δώματα· νεόμενος δ’
εἰς ἀγόρους ἀυτεῖ
τὰν κερόεσσαν ἔχειν 725
χρυσεόμαλλον κατὰ δῶμα ποίμναν.
27
Csapo (2003) 71–3; (2009).
146 Naomi Weiss

From beneath its tender {mother in the Argive{ mountains, as the
rumour remains among grey-haired tales, once Pan, guardian of fields,
blowing on well-fitted reeds sweet-strained music, brought forth the
golden-fleeced lamb. And standing on a stone platform, the herald cries
out: ‘Make your way to the agora, to the agora, Myceneans, to see the
blessed royals’ prodigies, {terrors{.’ {And choruses{ began to honor the
house of the Atreidae.
Altars of beaten gold were spread, and through the Argives’ city the fire
on the altar was gleaming. And the sound of the lōtos pipe resounds,
most beautiful, the Muses’ servant, and lovely songs were swelling forth,
{in praise{ of the golden fleece of Thyestes: for having persuaded the dear
wife of Atreus in secret union, he carries the portent out to his house. And
coming into the agora he shouts that he has the horned, golden-woolled
sheep at his home.
In the opening strophe the description of Pan’s μουσική, along with
its simultaneous enactment in the theatre, makes the initial depiction
of pastoral simplicity particularly vivid: as the chorus sings of the god
playing his syrinx, the audience would hear the tune of the aulos.28
The image of the syrinx with its ‘sweet-strained music’ (μοῦσαν
ἡδύθροον, 703) shapes the audience’s reception of the sound of the
aulos, so that this can momentarily represent for them Pan’s piping.
The metonym of ‘well-fitted reeds’ (εὐαρμόστοις . . . καλάμοις, 702)
contributes to this merging of described and performed μουσική,
since κάλαμοι were associated with the aulos as well as with the
syrinx.29 This mimetic effect helps to transport us to a peaceful, bucolic
scene, far from the imminent bloodshed of the dramatic present.
The crossover between described and performed μουσική con-
tinues into the antistrophe. As in the preceding strophe, the chorus
places particular emphasis on instrumental noise—now that of
the lōtos pipe, a name often given to the aulos in Euripidean tragedy:
‘the sound of the lōtos pipe resounds, most beautiful, the Muses’
servant’ (λωτὸς δὲ φθόγγον κελάδει / κάλλιστον, Μουσᾶν θεράπων,
716–17). This description of the aulos comes in exact responsion
with the lines in the strophe describing the syrinx, further encouraging
28
Gagné and Hopman (2013) 8 also note this mimetic effect: ‘When [the chorus]
sings of Pan blowing sweet music in his harmonious pipes, a direct link is established
between the sound of the poetic reeds and the sound of the aulos in the orchestra.’
29
Syrinx associated with κάλαμοι (instead of δονάκες): Eur. IA 577, 1038; El. 702;
IT 1125–7; Ar. fr. 719 (καλαμίνην σύριγγα). Aulos and κάλαμοι: Theophrastus 4.6; Ar.
fr. 144; Theoc. Id. 5.6–7; Ath. 4.78, 4.80.5–6.

a sense of merging between the two instruments.30 We can assume that
the tune of the auletic accompaniment would be similar (or even
identical) for both the strophe and antistrophe, yet its sound is
imagined as two separate but closely related instruments within the
same song. The transition from there to here, from country to city,
mountains to Argos, is thus enacted musically through the transition
from the rustic syrinx to the urban aulos. It also brings us from then to
now, from the carefree, pastoral past to the more disturbing dramatic
present; this shift may have a particularly powerful effect now that the
described sound matches the aulos-playing in the here and now of the
performance. The movement back towards the adultery and murder in
Argos becomes clear through the language of the song as well, as Csapo
has shown: the reality of the dramatic situation first intrudes through
the prominently positioned Θυέστου in line 720 (Thyestes has the
fleece, not Atreus); the chorus then explain how Thyestes stole the
fleece after luring Atreus’ wife to bed.31
It is surely no coincidence that the syrinx makes this unprecedent-
edly marked appearance at this point in Euripides’ career, when he
starts to engage with the so-called ‘New Music’. It is also not surprising
that it should appear in connection with the aulos, which is at the
centre of much of the discourse surrounding new musical trends in
Athens in the late fifth and early fourth centuries. One of the charac-
teristics of the aulos that is most emphasized in Greek texts, whether
positively or negatively, is its mimetic flexibility: Pindar, for example,
repeatedly calls it or its tune ‘all-sounding’ (πάμφωνος), thereby refer-
ring not only to the loudness of its music but also to its perceived ability
to represent anything;32 in Pythian 12, Athena invents the ‘many-
headed nomos’ (κεφαλᾶν πολλᾶν νόμον, 23) for the aulos so that she
can imitate (μιμήσαιτ’, 21) Euryale’s dying wail.33 According to the
conservative criticism of new styles of μουσική in Plato’s Laws, imita-
tion of different types of sounds in general seems to have been a
popular trend among musicians of this period: the Athenian Stranger
complains that the Muses ‘would never combine the cries of beasts and
of humans and of instruments and all kinds of noises into the same
30
For further discussion of the effects created by stanzaic responsion, see the
pieces by Phillips and Thomas in this volume.
31
Csapo (2009) 98.
32
Pind. Ol. 7.12; Pyth. 12.19–21; Isth. 5.27. On the mimetic powers of the aulos, see
esp. Barker (1984) 51; Wilson (1999) 87–93.
33
On the possible role of musical mimesis in Pythian 12, see esp. Phillips (2013).
148 Naomi Weiss

piece, as a way to represent one thing’ (θηρίων φωνὰς καὶ ἀνθρώπων καὶ
ὀργάνων καὶ πάντας ψόφους εἰς ταὐτὸν οὐκ ἄν ποτε συνθεῖεν, ὡς ἕν τι
μιμούμεναι, 669c8–d1). The famously ‘mimetic’ aulos could most easily
appear to imitate all these things: from bestial cries, like those of the
serpent in performances of the Pythikos nomos or of Scylla, whom,
according to Aristotle, bad auletes would enact by physically dragging
the chorus leader around, to other instruments, of which the syrinx
would presumably be one of the most straightforward to represent—
conceptually, acoustically, and visually.34
The mimetic powers of the aulos are linked to its reputation (at
least from the late fifth century onward) for complex modulation, an
association that is neatly expressed by Socrates’ condemnation of it as
‘the most many stringed’ (πολυχορδότατος) instrument in Plato’s
Republic (399c7–d9):
οὐκ ἄρα, ἦν δ’ ἐγώ, πολυχορδίας γε οὐδὲ παναρμονίου ἡμῖν δεήσει ἐν ταῖς
ὠιδαῖς τε καὶ μέλεσιν.
οὔ μοι, ἔφη, φαίνεται.
τριγώνων ἄρα καὶ πηκτίδων καὶ πάντων ὀργάνων ὅσα πολύχορδα καὶ
πολυαρμόνια, δημιουργοὺς οὐ θρέψομεν.
οὐ φαινόμεθα.
τί δέ; αὐλοποιοὺς ἢ αὐλητὰς παραδέξηι εἰς τὴν πόλιν; ἢ οὐ τοῦτο
πολυχορδότατον, καὶ αὐτὰ τὰ παναρμόνια αὐλοῦ τυγχάνει ὄντα μίμημα;
δῆλα δή, ἦ δ’ ὅς.
λύρα δή σοι, ἦν δ’ ἐγώ, καὶ κιθάρα λείπεται κατὰ πόλιν χρήσιμα· καὶ αὖ κατ’
ἀγροὺς τοῖς νομεῦσι σύριγξ ἄν τις εἴη.
‘Well then, we won’t need many-stringed and panharmonic [instruments] in
our songs and melodies.’
‘No,’ he said, ‘it doesn’t seem so to me.’
‘Then we won’t support craftsmen of trigōnoi [triangular harps] or pēktides
[a harp played with the fingers] or any [of the instruments] that are many-
stringed and polyharmonic.’
‘Apparently not.’
‘What then? Will you admit aulos makers and aulos players into the city? Or
isn’t this the most many-stringed [instrument], and aren’t the panharmonic
ones themselves actually an imitation of the aulos?’
‘Clearly,’ he said.
‘You have left,’ I said, ‘the lyre and the kithara as things that are useful in the
city; and in the fields the shepherds would have some sort of syrinx.’
34
Arist. Poet. 1461b31–2. For a description of the Pythikos nomos, see Poll. 4.84.

Though πολυχορδότατος primarily refers to the versatility of the
aulos in shifting between different scale forms, the application of
an adjective more suited to the kithara also suggests its flexibility
in assuming different instrumental identities. It even seems to
encourage imitation as well as being imitative itself, a trait which
makes it appear doubly threatening to the city they are constructing:
other ‘panharmonic’ instruments (presumably meaning those which
can flit between different ἁρμονίαι) are in turn an imitation (μίμημα,
399d5) of the aulos.35 Interestingly, Socrates then states that the only
type of pipe to be permitted in the city is a type of syrinx for
shepherds to play in the countryside (399d7–9). The singular σύριγξ
here may suggest only a simple pipe rather than the plural panpipes,
but the contrast with the aulos effectively demonstrates the perceived
closeness and distance between the two instruments rather as the
second stasimon of Euripides’ Electra does, but in the reverse order:
the rustic syrinx is to replace the urban aulos as another type of
pipe(s), but one which seems far less dangerous.
The syrinx therefore enters the theatre through the sound of the
aulos, exploiting and showcasing the latter’s perceived mimetic ability.
Indeed, the syrinx almost always appears in Euripides’ tragedies
either along with the aulos or in such a way that the aulos is expected
or at least strongly suggested instead. In Troades, for example, when
Hecuba performs her own version of a parodos with her ‘chorus-less
cries’ (ἄτας . . . ἀχορεύτους, 121), she vividly remembers the arrival of
the Greek ships at Troy, accompanied by the sound of auloi and
syrinxes (Eur. Tro. 122–30):
πρῶιραι ναῶν, ὠκείαις
Ἴλιον ἱερὰν αἳ κώπαις
δι’ ἅλα πορφυροειδῆ καὶ
λιμένας Ἑλλάδος εὐόρμους 125
αὐλῶν παιᾶνι στυγνῶι
συρίγγων τ’ εὐφθόγγων φωνᾶι
βαίνουσαι {πλεκτὰν Αἰγύπτου
παιδείαν ἐξηρτήσασθ’{,
αἰαῖ, Τροίας ἐν κόλποις . . . . 130
35
Cf. Wilson (1999) 93 on the aulos as ‘the model of musical mutiplicity’. On the
‘aulization’ of the kithara, especially from the late fifth century onward, see Power
(2013) 243–4.
150 Naomi Weiss

Prows of ships, which with swift oars to holy Ilium over the dark purple
sea and the fair harbours of Hellas, to the hateful paean of auloi and
the voice of fine-sounding syrinxes, travelling, {you hung the twisted
handiwork of Egypt{—aiai—in the bays of Troy. . . .
It is likely that the aulete, even if not yet present on stage, would at
this point already be playing in accompaniment to Hecuba’s ana-
paests, particularly given the paradoxically choral character of her
monody, stressing the absence of χορεία in the aftermath of Troy’s
destruction: not only does she present her song as one which replaces
a parodos, but her address to the ship’s prows is more typical
of a choral ode (the first stasimon of Euripides’ Electra, which
I mentioned above, similarly begins with a hanging apostrophe to
the ‘famous ships’ (κλειναὶ νᾶες, 432) heading to Troy).36 When she
sings of the Greeks’ musical accompaniment, ‘the hateful paean of
auloi and the voice of fine-sounding syrinxes’ (αὐλῶν παιᾶνι στυγνῶι /
συρίγγων τ’ εὐφθόγγων φωνᾶι, 126–7), the sound of the aulos in the
theatre would therefore represent that of both instruments, produ-
cing a particularly vivid sound picture of the invasion. Even here,
however, we can see a difference in the conceptualization of the two
instruments’ sound: the more threatening aulos against the ‘fine-
sounding’ syrinx.
Even when a syrinx is described in a song without any explicit
mention of an aulos, the sound of the auletic accompaniment in
performance might introduce a similarly foreboding tone, producing
a dissonance as well as a crossover between described and performed
sound. In Ion, one of the numerous accounts of Creusa’s rape and Ion’s
birth is brought to life when the chorus sings of the daughters of
Aglauros dancing on the Acropolis, singing to the accompaniment of
Pan’s syrinx (Eur. Ion 491–506):37
ὦ Πανὸς θακήματα καὶ
παραυλίζουσα πέτρα
μυχώδεσι Μακραῖς,
ἵνα χοροὺς στείβουσι ποδοῖν 495
Ἀγλαύρου κόραι τρίγονοι
36
Cf. Eur. Hel. 1451–64; IT 1123–37. On the apparent postponement of χορεία in
Troades, see Weiss (2018) 106–10.
37
On the repeated accounts of this scene, see Weiss (2008) 41–5; (2012) 39–40.
Pan is also represented as a chorus leader as he plays on his syrinx in Hom.h.Pan
14–26.
στάδια χλοερὰ πρὸ Παλλάδος

ναῶν συρίγγων
ὑπ’ αἰόλας ἰαχᾶς
{ὕμνων{ ὅτ’ ἀναλίοις 500
συρίζεις, ὦ Πάν,
τοῖσι σοῖς ἐν ἄντροις,
ἵνα τεκοῦσά τις
παρθένος μελέα βρέφος
Φοίβωι πτανοῖς ἐξόρισεν
θοίναν θηρσί τε φονίαν 505
δαῖτα, πικρῶν γάμων ὕβριν.
O resting place of Pan and rock that lies near the cavernous Long Rocks,
where three daughters of Aglauros step [in] choruses, over the grassy
courses before the temple of Pallas, to the accompaniment of the
shimmering cry of the syrinx, {singing{, when you play your syrinx
in your sunless caves, Pan, where a wretched maiden bore a child to
Phoebus and cast it out as a meal for birds and bloody feast for wild
beasts, the violent fruit of her bitter union.
While it sings of this other dancing group, the dramatic chorus is
itself singing and dancing to the sound of another pipe, the aulos, with
the result that the audience is invited to merge the two performances,
to see (and hear) one as the other; the syrinx in the song thus becomes
the aulos in the theatre (and vice-versa). The audience is encouraged to
perceive the aulos as a syrinx, and hear in its sound Pan’s piping, so
that they can be vividly transported to the scene that is such a crucial
moment of Creusa’s (and Athens’) past. The description of Pan’s syrinx-
playing heightens the scene’s rusticity, but the actual sound of the aulos
may underscore quite how profoundly disturbing it is. This impression
is created verbally by the long enjambment, which begins with such an
apparently peaceful, bucolic image, finally ending in line 506 with the
word ὕβρις, which, though referring to the abandoned baby, strongly
suggests the violence of the rape itself.
The aulos is doubly suggested in this passage of Ion, not only
through the crossover of the syrinx in the song and the aulos in the
theatre, but also through the suggestive nature of the chorus’ lan-
guage. It describes the syrinx’s sound as αἰόλος, which suggests a
synaesthetic mix of sight and sound, meaning both ‘nimble, quick-
moving, changing’ and ‘glittering, shimmering’. In this respect it is
very like the adjective ποικίλος and noun ποικιλία, which, as Pauline
LeVen has shown, capture a combination of the visual and aural
152 Naomi Weiss

effects of μουσική.38 As Csapo in particular has demonstrated, in
critical discourse concerning new musical trends ποικιλία not only
refers to musical complexity but also becomes a sociopolitically
charged term, as ‘a symbol for plurality, changeability, innovation,
openness, liberation, inclusiveness and mixing’.39 Like poly-compounds
(such as πολυχορδότατος), ποικιλία seems to have been especially
associated with the aulos, as the famous fragment of Pratinas’ satyr
play suggests: there the chorus describes its tune as the ‘breath of a
spotted [ποικίλος] toad’ (τὸν φρυνεοῦ ποικίλου πνοὰν ἔχοντα, fr. 708
PMG 10)—a description which may also refer to the aulete’s bulging
cheeks, the result of his circular breathing.40 In the performance of
this choral song ποικίλος could have a further metatheatrical refer-
ence, possibly pointing to the aulete’s costume with its typical decor-
ation of πάσματα, which seem to have resembled large sequins. The
adjective αἰόλος and related words also seem to have been among the
buzzwords of the new musical trends in the fifth century: Timothy
Power has recently suggested that when Hermes in Sophocles’ Ich-
neutai is said to be ‘lifted up’ by the lyre’s αἰόλισμα (ξύμφωνον ἐξαίρει
γὰρ αὐτὸν αἰόλισμα τῆς λύρας, TGrF F 314.327), this term alludes to
the increasing complexity of new styles of kithara music.41 So in
Euripides’ Ion, though the chorus sings of the rustic syrinx, what
the audience would hear and see was the innovative aulos, a much
more apt candidate for the similar-sounding adjective αἰόλος, both in
terms of its tune and broader conceptualization, and perhaps also as a
result of the aulete’s elaborate costume.
In both Iphigenia in Tauris and Helen, the chorus mentions the
syrinx while describing the female protagonist’s escape by sea, when
we would expect the musical accompaniment to be the aulos instead:
this, not the syrinx, was typically played for the rowers of a trireme; it
is also part of the dithyrambic imaginary, linked with dolphins,
38
LeVen (2013). On the semantics of ποικιλία and related words, particularly in
relation to the composition of a song, see Nagy (1996), esp. 59–66.
39
Csapo and Wilson (2009) 291–2; cf. Csapo (2004), esp. 227–30; LeVen (2013)
240–1; (2014) 101–3.
40
This fragment suggests that ποικίλος/ποικιλία could already be a charged term
even in the late sixth or early fifth century, to which it is traditionally dated. Some have
proposed a later date (and therefore another, otherwise unattested Pratinas as its poet)
on account of the fragment’s highly metamusical performance criticism, but there
seems to be little reason to doubt its authenticity as belonging to Pratinas of Phlious:
for a review of this debate, see Griffith (2013) 273 n. 57.
41
Power forthcoming.

Dionysus, and maritime travel and trade.42 In Helen, the syrinx
appears in the third stasimon, when, after imagining the sea journey
of Helen, its former χορηγός, back to Sparta, the chorus expresses its
wish to fly away with strongly choreographic and musical language
(Eur. Hel. 1478–94):
δι’ αἰθέρος εἴθε ποτανοὶ
γενοίμεθ’ ὅπαι Λιβύας
οἰωνῶν στιχάδες 1480
ὄμβρον χειμέριον λιποῦ-
σαι νίσονται πρεσβυτάτου
σύριγγι πειθόμεναι
ποιμένος, ἄβροχά θ’ ὃς
πεδία καρποφόρα τε γᾶς 1485
ἐπιπετόμενος ἰαχεῖ.
ὦ πταναὶ δολιχαύχενες,
σύννομοι νεφέων δρόμωι,
βᾶτε Πλειάδας ὑπὸ μέσας
Ὠρίωνά τ’ ἐννύχιον, 1490
καρύξατ’ ἀγγελίαν
Εὐρώταν ἐφεζόμεναι,
Μενέλεως ὅτι Δαρδάνου
πόλιν ἑλὼν δόμον ἥξει.
If only we could be flying through the air, where the rows of birds from
Libya go, leaving the wintry rain, obeying the syrinx of the eldest, the
shepherd who, winging his way over the unwetted and crop-bearing
plains of the earth, cries out. O long-necked winged creatures, partners
of the clouds in [your] racing, go beneath the Pleiades in midcourse and
Orion in the night. Announce the news as you land by the Eurotas, that
Menelaus, having taken the city of Dardanus, will come home.
The chorus longs to be transformed into Libyan birds (most likely
cranes) migrating to Greece, following its syrinx-playing shepherd
leader in ‘rows’ (στιχάδες, 1480), which might evoke the typical
v-formation in which these birds were known to fly, and could also
be quite easily represented in the chorus’ own choreography.43 It is
even possible, as Deborah Steiner suggests, that the chorus’ wish to be
42
On this network of associations, see Csapo (2003); Kowalzig (2013a).
43
Their identification as cranes is suggested by the migration described here from
Northern Africa, where eastern European cranes tend to winter: see Arnott (2007) 80.
Their v-formation is noted in Plut. Mor. 967b–c, 979a; Ael. NA 3.13; Cic. De nat. deor.
2.49.125; Phil. Her. 11.4.
154 Naomi Weiss

these birds could evoke the Athenian γέρανος (crane) dance that
Theseus was said to have invented and, according to Callimachus,
Plutarch, and Pollux, was performed at Delos; sources differ as to
whether the dance’s choreography was circular or linear.44 What is
important here, however, is that the Athenian audience might readily
associate cranes with χορεία in general, and so link the chorus’
description of its movement with the dancing being performed in
the theatre. The choral identification of these birds is further made
evident through the direction in line 1489 that they should fly beneath
the Pleiades, the archetypal star chorus.45
Given the depiction of Helen in her role as chorus leader in the
previous antistrophe (1465–77) and her ship’s journey to Greece
described in the opening strophe (1452–64), the syrinx-playing
crane whom the chorus wishes to follow to Sparta here must repre-
sent Helen herself, their absent χορηγός who has left them behind.46
As we have seen, however, the image of this instrumentalist can
simultaneously have a metatheatrical reference and be linked to the
aulos player accompanying the chorus’ dance: the acoustic image of
the syrinx representing the cranes’ cry would merge with the tune of
the aulos being played in the theatre, creating a particularly vivid
sound picture.47 The aulete could therefore visually assume a role
similar to that of the χορηγός whom the chorus describe, adding to
the epiphanic effect of their performance: Helen can be imagined to
be there with them, represented by their aulos-playing leader.
In Iphigenia in Tauris, the syrinx appears twice, both times in the
context of maritime travel: in the first stasimon, when the chorus
vividly imagines the journey of the ‘strangers’ (Orestes and Pylades)
from Argos to Tauris; then in the second stasimon, when, as in Helen,
it describes the female protagonist’s escape by sea, though this turns
out to be a purely imaginative journey that will not in fact take place.
When simply read as text, the first occurrence of a syrinx-related
44
Steiner (2011) 314–15. The γέρανος dance is described in Call. H.4.310–13; Plut.
Thes. 21; Poll. 4.101; cf. also Luc. Orch. 34.
45
Cf. Padel (1974) 237; Steiner (2011) 316–17. On the Pleiades as a chorus of stars,
see Csapo (2008) 266–7.
46
On Helen’s gradual departure from the chorus as she heads toward the choral
role she will soon resume in Sparta, see Murnaghan (2013).
47
See also Steiner (2011) 311–12 on the merging of the aulete with the syrinx-
playing crane-leader here.

word may initially seem simply to indicate the sound of the ship’s
steering oars as they cut through the water (Eur. IT 422–38):
πῶς τὰς συνδρομάδας πέτρας,
πῶς Φινεΐδας {ἀύ-
πνους{ ἀκτὰς ἐπέρα-
σαν, παρ’ ἅλιον αἰγιαλὸν ἐπ’ Ἀμφιτρί- 425
τας ῥοθίωι δραμόντες,
ὅπου πεντήκοντα κορᾶν
Νηρήιδων ⟨— ˘⟩ χοροὶ
μέλπουσιν ἐγκύκλιοι,
πλησιστίοισι πνοαῖς, 430
συριζόντων κατὰ πρύμναν
εὐναίων πηδαλίων,
αὔραισιν νοτίαις
ἢ πνεύμασι Ζεφύρου,
τὰν πολυόρνιθον ἐπ’ αἶ-
αν, λευκὰν ἀκτάν, Ἀχιλῆ- 435
ος δρόμους καλλισταδίους,
ἄξεινον κατὰ πόντον;
How did they pass the running together rocks, how the {unsleeping{
shores of Phineus, running along the sea-shore on Amphitrite’s surf,
where choruses of fifty Nereid maidens sing in a circle, as the guiding
steering-oars pipe forth at the stern with sail-filling gusts, with southerly
breezes or Zephyr’s breaths, to the land with many birds, the white
shore, Achilles’ fine running courses, across the unhospitable sea?
When performed in the theatre, however, the ‘piping’ or ‘whistling’
(συριζόντων) of the oars would simultaneously be produced by the
auletic accompaniment. The participle further suggests the aulos as a
result of the preceding description of the chorus of fifty Nereids
singing in a circle, since this evokes the performance of a dithyramb,
which was likewise danced by fifty choreuts in a circular formation to
the sound of the aulos;48 the image also combines with the χορεία
being simultaneously performed in the theatre. The emphasis on
breath in these lines (πνοαῖς in 430, πνεύμασι in 434) suggests the
blowing of pipes (those of the aulos or syrinx) as much as it does gusts
of wind.
48
On the circular formation of the dithyramb, see D’Angour (1997). On the
association of Nereids (especially fifty of them) with the dithyramb, see Csapo (2003).
156 Naomi Weiss

In the second stasimon of Iphigenia in Tauris the syrinx also
appears as an accompaniment to naval travel, this time alongside
the kithara in a multilayered soundscape. Singing of Iphigenia’s
escape by sea, the chorus imagines her on a ship with Pan and Apollo
as divine and musical escorts (Eur. IT 1123–33):
καὶ σὲ μέν, πότνι’, Ἀργεία

πεντηκόντερος οἶκον ἄξει·
συρίζων θ’ ὁ κηρόδετος 1125
Πανὸς οὐρείου κάλαμος
κώπαις ἐπιθωύξει,
ὁ Φοῖβός θ’ ὁ μάντις ἔχων
κέλαδον ἑπτατόνου λύρας
ἀείδων ἄξει λιπαρὰν 1130
εὖ σ’ Ἀθηναίων ἐπὶ γᾶν.
{ἐμὲ δ’ αὐτοῦ λιποῦσα
βήσηι ῥοθίοις πλάταις. . . .
And you, lady, an Argive ship with fifty rowers shall bring you home,
and the wax-bound reed of Pan, the mountain god, will blow and shout
out to the oars, while Phoebus the prophet, holding the noisy seven-
stringed lyre, will sing and lead you safely to the gleaming land of Athens.
{But me you shall leave here and make your way with splashing oars . . . .
Like the fifty Nereids in the previous ode, the fifty-oared ship
encourages us to equate the oars with the fifty choreuts dancing a
dithyramb, yet it is not an aulete but Pan with his syrinx who
accompanies the rowing—though of course in the performance of
their song it would in fact be the aulos that would ‘shout out’
(ἐπιθωύξει, 1127) to the choreuts, who are assimilated with the rowers
as well as the oars. Through the crossover between this image of
dithyrambic μουσική, with its indirect suggestion of the aulos, and
the chorus’ own performance to the accompaniment of the aulos
in the theatre, it seems to enact Iphigenia’s journey itself. In this
play such a mimetic performance is ironic, serving as a stark contrast
with the chorus’ own journey to Tauris on a slave ship, which it
laments in the previous antistrophe (1106–22), though it perhaps also
looks forward to the women’s own departure for Greece at the end of
the drama, following Athena’s instructions at 1468–9. Unlike Helen,
Iphigenia is unsuccessful in her attempt to escape with Orestes and
Pylades, and will in fact end up in Brauron rather than Athens
(1462–7), so the chorus’ performance here also ironically represents

a fiction, inducing the audience temporarily to suspend disbelief and
imagine a journey that will not actually take place.
Like the syrinx, Apollo’s ‘seven-stringed lyre’ (ἑπτατόνου λύρας,
1129) belongs to the imagined μουσική of the song. The audience,
however, would nonetheless hear loud musical noise (κέλαδον, 1129)
as a result of the auletic accompaniment, which would merge with
and represent the sound the chorus describes, even though the lyre
seems so different—acoustically, visually, and conceptually—from
the reed pipes. We have already seen that the aulos was long associ-
ated with mimetic flexibility, and that Socrates’ description of it as
‘most many-stringed’ (πολυχορδότατος) in Plato’s Republic may sug-
gest that it could be imagined even to imitate a stringed instrument
like the lyre. The dithyrambist Melanippides may in particular have
exploited its mimetic flexibility, if his Marsyas narrated the musical
contest between Apollo on the lyre and Marsyas on the aulos: rather
than this involving an actual kithara extraordinarily being played on
stage, as John Boardman suggests, the aulos itself, along with the
chorus’ song, could have represented the god’s music.49 There may be
a similar mimetic effect at the start of the third stasimon of Euripides’
Iphigenia in Aulis, when the chorus sings of the aulos, kithara, and
syrinxes playing at the wedding of Peleus and Thetis (1036–9): here
the sound of the aulos (as well as the chorus’ own song and dance)
works metonymically as a (re)enactment of all three instruments,
thereby transporting the audience back to the wedding celebrations
49
Boardman (1956). His argument that the dithyramb included an enactment of
Marysas’ transformation from aulete to kitharode relies on a series of late fifth-century
vase paintings that show Marsyas playing Apollo’s lyre; cf. Csapo (2004) 213. For a
more cautious approach to Melanippides’ Marsyas, see Power (2013) 240–2: he points
out that there is little evidence for the inclusion of lyre music in the dithyramb, but
nonetheless suggests that ‘Melanippides, if he did narrate the “conversion” of the
champion of the aulos, [could] already have encoded in his dithyramb a critique of
contemporary kitharoidia, in particular its increasing flirtations with Dionysiac music
(the aulos), themes, and histrionics’ (242). If the kithara was not typically brought on
stage in the performance of dithyrambs, then the aulos might have played a similarly
mimetic role in Melanippides’ version of the story of the lyre singer Linos, though we
do not know if this was in fact a dithyramb: Melanippides was most associated with
the dithyrambic genre (e.g. Xen. Mem. 1.4.3; Suda M 454), but in Ps.-Plutarch’s De
musica, just after he is called ‘the composer of dithyrambs’ (τὸν τῶν διθυράμβων
ποιητήν, 1141d2), comes the fragment from Pherecrates’ Cheiron in which Music
complains that he ‘loosened me up with his twelve strings’ (χαλαρωτέραν τ’ ἐποίησε
χορδαῖς δώδεκα, 1141e3), thereby seeming to refer expicitly to his kithara music
instead: see Barker (1984) 93.
158 Naomi Weiss

it describes. At the same time, the chorus thus seems to perform its
own, horribly ironic ὑμέναιος for Iphigenia and Achilles, whom the
audience knows will never be married.50
This sort of musical experimentation in Iphigenia in Tauris and
Iphigenia in Aulis suggests that Euripides was particularly interested
in creating mimetic effects through the aulos toward the end of the fifth
century. But the third stasimon of Alcestis, which was produced in 438
BC, demonstrates that he was doing so much earlier in his career as well,
even if most of the surviving examples come from around 420 onward.
The chorus begins this song by remembering Apollo’s stay at Pherae,
when he exchanged his kithara for the syrinx (Eur. Alc. 569–87):
ὦ πολύξεινος καὶ ἐλευθέρου ἀνδρὸς ἀεί ποτ’
οἶκος,
σέ τοι καὶ ὁ Πύθιος εὐλύρας Ἀπόλλων 570
ἠξίωσε ναίειν,
ἔτλα δὲ σοῖσι μηλονόμας
ἐν νομοῖς γενέσθαι,
δοχμιᾶν διὰ κλειτύων 575
βοσκήμασι σοῖσι συρίζων
ποιμνίτας ὑμεναίους.
σὺν δ’ ἐποιμαίνοντο χαρᾶι μελέων βαλιαί τε
λύγκες,
ἔβα δὲ λιποῦσ’ Ὄθρυος νάπαν λεόντων 580
ἁ δαφοινὸς ἴλα·
χόρευσε δ’ ἀμφὶ σὰν κιθάραν,
Φοῖβε, ποικιλόθριξ
νεβρὸς ὑψικόμων πέραν 585
βαίνουσ’ ἐλατᾶν σφυρῶι κούφωι,
χαίρουσ’ εὔφρονι μολπᾶι.
O house of a hospitable and ever generous man, even Pythian Apollo of
the lovely lyre deigned to dwell in you and submitted to become a
shepherd in your pastures, playing on his syrinx wedding songs for your
herds on the slanting hillsides.
And, in joy at his songs, both spotted lynxes began to be shepherded,
and, leaving the vale of Othrys, a tawny troop of lions came, and the
dappled fawn, stepping beyond the tall fir trees, danced to your kithara
with its light foot, rejoicing in the joyful song.
50
See Weiss forthcoming for a full discussion of the merging of performed and
imagined μουσική in this song.

Though the character Apollo is no longer on stage, he seems to be
present through the crossover of his piping as described by the chorus
in the opening strophe and that of the aulete heard by the audience.
As in the second stasimon of Iphigenia in Tauris, however, this sound
picture is multilayered, including the kithara as well as the syrinx: in
the antistrophe Apollo’s characteristic instrument reappears, as the
chorus sings of him as an Orpheus figure, charming wild beasts with
his music-making, to which they dance in a chorus (χόρευσε, 583).
The syrinx has turned into the kithara, but the aulos, to which the
dramatic chorus is dancing, represents both, producing the epiphanic
effect of Apollo’s presence in the theatre.
The third stasimon of Alcestis demonstrates that this sort of instru-
mental mimesis in tragedy was not confined to the so-called ‘New
Music’ of the late fifth century. It may not be a coincidence, however,
that this passage appears in a play which, as has often been noted,
contains many satyric traits, from its position as the fourth drama in
Euripides’ tetralogy to the scene of Heracles’ drunkenness.51 Since, as
Mark Griffith has recently shown, metamusicality seems to have been
a hallmark of satyr drama, this early experimentation with the rep-
resentation of different instruments’ sounds in a tragedy may there-
fore also be a result of Euripides’ mixing of the two genres.52
Unfortunately, however, too few of Euripides’ plays survive from
the 430s for us to know the extent to which Alcestis was unusual in
sharing these elements with satyr play.
Nonetheless, the pattern that emerges from the extant Euripidean
tragedies, with the exception of Alcestis, is one of increasing experi-
mentation with different musical effects in the latter part of his career.
Almost all the references to the syrinx in surviving tragedy date from
around 420 or later, and they all make use of the aulos being played in
the theatre, drawing on its conceptualization as a highly mimetic
instrument. Such exploitation of the two instruments’ mimetic rela-
tionship reaches its peak with a rather confusing few lines in the first
51
On Alcestis as ‘satyric’ or ‘prosatyric’, see esp. Marshall (2000); Slater (2005);
Shaw (2014) 94–105. For a more sceptical approach to this sort of generic classifica-
tion, see Mastronarde (2010) 56–7.
52
Griffith (2013); cf. Power forthcoming. This sort of cross-fertilization between
satyr drama and tragedy may have resulted partly from the generic boundaries
between them being more fluid than we tend to assume: Mastronarde (2010) 57
points out that such elements would not necessarily have been recognized as ‘inher-
ently satyric and non-tragic’ by fifth-century audiences.
160 Naomi Weiss

stasimon of one of Euripides’ very last plays, Iphigenia in Aulis, which
was produced posthumously, probably in 405 BC.53 After some gno-
mic speculation regarding moderation and virtue in love (543–72),
the chorus describes how Paris arrived at the Judgement scene on
Mount Ida (573–81), and then pictures him standing before Helen’s
palace, sparking the love between them and so setting the Trojan War
in motion (582–6). The depiction of destructive love and strife in this
latter scene (intensified through the wordplay of ἔρως and ἔρις)
contrasts markedly not only with the statements in praise of
restrained love in the first strophic pair, but also with the preceding
image of pastoral calm and simplicity, which includes the particularly
vivid detail of Paris playing on his syrinx (573–8):
{ἔμολες, ὦ Πάρις, ἧιτε σύ γε{
βουκόλος ἀργενναῖς ἐτράφης
Ἰδαίαις παρὰ μόσχοις, 575
βάρβαρα συρίζων, Φρυγίων
αὐλῶν Οὐλύμπου καλάμοις
μιμήματα {πνέων{.
{‘You came, Paris, to the place where{ you were reared as a herdsman
among the shining white heifers of Mount Ida, piping foreign tunes on
the syrinx, {breathing{ on the reeds imitations (mimēmata) of the
Phrygian auloi of Olympus.’
Translations of lines 576–8 vary. Andrew Barker suggests ‘breath-
ing imitations of Olympus on the reeds of Phrygian auloi’, but notes
that ‘it is not clear which instrument Paris is playing here’.54 David
Kovacs understands the objective genitive going with μιμήματα to be
the auloi rather than Olympus, and so translates these lines as
‘imitating upon your reed pipe / the Phrygian aulos of Olympus’.55
James Morwood and Chris Collard, in their commentary on Iphige-
nia in Aulis, give the translation ‘playing your barbarian tunes and, as
you breathed on the reeds, imitating the Phrygian pipe of Olympus’,
but then note that the literal translation is ‘breathing imitations of
Olympus on the reeds of your Phrygian pipe’.56 Any precise transla-
tion of these lines is made even more difficult due to the fact that we
53
According the scholion on Ar. Ran. 66–7, Iphigenia in Aulis was produced
alongside Bacchae and Alcmeon in Corinth.
54 55
Barker (1984) 92. Kovacs (2002) 223.
56
Collard and Morwood (2017).

could also understand Olympus not as the semi-legendary Phrygian
musician associated with the invention of the aulos, but as Mount
Olympus in Phrygia, in which case the translation would be ‘breath-
ing imitations of Phrygian auloi on the reeds of [Mount] Olympus’.
Moreover, not only is the word order problematic, but, as Barker
points out, it is unclear whether Paris is playing the syrinx or the aulos
here.57 Since the syrinx is traditionally associated with herdsmen
(though not necessarily Paris), we expect the participle συρίζων in
line 576 to have the specific meaning of ‘playing the syrinx’, yet here it
seems to refer to a performance on the aulos as well.
But to focus our efforts on the sole, precise meaning of these lines
would be to miss the metamusical play at work in this song. As a
result of the other examples of instrumental mimesis that I have
discussed here, we can see that Paris is playing both instruments,
for while the chorus sings of him imitating the aulos with his syrinx,
the audience hears the aulos imitating the syrinx; words and music
combine to create a doubling effect whereby both instruments can be
experienced at once. As in the first stasimon of Iphigenia in Tauris,
the verb συρίζω as it is used here encompasses the playing of both the
syrinx and the aulos, rather as φόρμιγξ and φορμίζω in the Homeric
epics and Hymns can refer to different types of lyre, not just the
phorminx.58 If auletes did occasionally use syrinx devices to raise the
pipes’ pitch in performances of tragedy, these lines might contain this
extra metamusical reference as well. All the possible translations of
these lines can therefore be correct, and their broadly suggestive
ambiguity is surely deliberate, exploiting the perceived intimacy
between the two instruments. Such an explicit pointer to instrumental
mimesis in this play from the end of Euripides’ life comes as the
climax of his musical experimentation with the sounds the aulos
could produce—or be imagined to produce.
This final example from Iphigenia in Aulis also demonstrates that
Euripides seldom produces these mimetic effects for their own sake,
nor just to display his virtuoso engagement with new musical trends.
Rather, such mimesis always has a particular dramatic function:
here the combination of the two instruments recreates both the
foreign, Phrygian soundscape of Mount Ida and the sense of bucolic
57
Barker (1984) 92 n. 198.
58
As in Hom.h.Merc., in which Hermes’ invention is variously called a chelys lyre,
phorminx, and kithara.
162 Naomi Weiss

innocence just before the stirrings of war; the shift to ἔρις in the
following lines then brings us back towards the situation at Aulis, and
also underscores the certainty of Iphigenia’s death, without which
there would be no Trojan War. But the dramatic effects of this sort of
interaction between words and music also result from it not always
being a strictly mimetic process. There remains a disconnect between
the two instruments even as they coincide, since the verbal descrip-
tion of sound does not fully transform the actual sound of the aulos
that the audience hears in the theatre, and consequently the auletic
accompaniment can add an ominous tone to the syrinx it is meant to
represent. We saw in the case of Electra the contrast between the two
sounds, both real and imagined, as the transition from the rustic
syrinx to the urban aulos with its more threatening associations
helps to return us to the murder at Argos, while in Ion I suggested
that the sound of the aulos could add a discomforting tone as the
chorus shifts from Pan’s pipe-playing to the rape of Creusa and her
exposure of her baby. In Iphigenia in Aulis the appearance of the
aulos once again works as a contrast to the pastoral, carefree syrinx,
heralding a shift back towards the problems of the dramatic present.
Whether or not the aulos is explicitly mentioned, then, its sound as an
accompaniment to song has a presencing effect, bringing the syrinx
and the scene in which the syrinx is played into the theatre, even
while it can problematize that imagined sound through its physical
representation. At the same time, the syrinx becomes a vehicle for
Euripides not just to spotlight the aulete’s performance, at a time
when this was becoming increasingly virtuoso, but also to exploit the
many strands of the auletic imaginary.59
59
On the professionalization of auletes through the fifth century, see esp. Wilson
(1999).
Lyric Atmospheres
Plato and Mimetic Evanescence
Anastasia-Erasmia Peponi
De la musique avant tout chose,

Et pour cela préfère l’ Impair
Plus vague et plus soluble dans l’air,
Sans rien en lui qui pèse ou qui pose.1
Nothing less than a manifesto about the eminent role of music in
poetry, these verses by Paul Verlaine, a key figure of French Symbol-
ism, embody in their intricate sonorities the very essence of what they
mean to convey: the ethereality and buoyancy of words. As the
symbolist movement imagined it, the musicality of language was to be
fully attained in flexible and supple metrical arrangements that dissolve
the compactness of verbal matter. Words, and with them signification,
become misty entities, their vagueness—here endorsed by Verlaine—to
be understood as similar to the blurred outlines of otherwise familiar
objects in a hazy landscape. The misty air that envelopes them softens
out their contours; although still recognizable, they become part of a
more unified, albeit evasive and uncanny, universe.
Such ideas may be representative of nineteenth-century concerns
about the musical potential of poetic language but, apart from their
1
Paul Verlaine, Art Poétique, published in 1884 in the collection Jadis et naguère.
Precisely because the sonic effects of the poem are so central to its ideas and atmos-
phere, it is almost untranslatable. For clarity’s sake, I offer here an utterly unambitious
translation into English: Music first and foremost, / Therefore prefer the metre odd, /
More soluble and vaguer in the air, / With nothing in it that weighs or halts.
164 Anastasia-Erasmia Peponi

distinctive ideological load, they are far from alien to issues that were
either latent or explicitly addressed in antiquity. For instance, how
unusual verbal constellations may enhance sonic effects and challenge
conventional modes of signification seems to have been a practice
pursued by the so-called New Music and its successive (and varying)
waves in the last part of the fifth century and the first part of the
fourth century BC.2 Later on, practices of blurring, indeed softening,
the sonic contours of individual words, thus enabling the emergence
of an expanded sonic environment, were well studied by ancient
literary critics like Dionysius of Halicarnassus.3
Hence, classicists have much to contribute to a broader historical
picture that, starting from such vigorous modern proclamations as,
for instance, Verlaine’s, can bring into the discussion relevant issues
debated in antiquity. In this chapter I should like to focus on one
particular set of issues. First, the idea that sonority and amplified
musicality in poetry create a mental environment in which plain
sense-making subsides, giving way to alternative modes of sense
perception. More specifically, I should like to discuss the view that
the sonic environment of poetry may operate as a dissolving medium,
somehow disintegrating the semantic concreteness of words while at
the same time emitting a certain atmosphere or mood that envelopes
the listener. Finally, I contend that in some of Plato’s dialogues
interesting versions of this broader issue are either openly addressed
or treated as an implicit struggle that results sometimes in negative,
while other times in particularly creative, responses. In either case
they illuminate neglected but exciting aspects of the philosopher’s
encounter with μουσική and the verbal arts.
ATMOSPHERES
Air is phonetically highlighted in Verlaine’s stanza. The word itself

appears at the end of the third line (plus soluble dans l’ air), but the
2
On the language of the New Music see for instance Csapo (2004), esp. 216–29;
Ford (2013) 313–31; LeVen (2014) 151–88.
3
I am referring primarily (but not exclusively) to Dionysius of Halicarnassus’
analysis of the polished style (γλαφυρὰ σύνθεσις) where he includes Sappho among
others. See D. H. De comp. 23. For a general approach to the emphasis put on the sonic
texture of language in Greek thought see for instance Porter (2010), esp. 365–404.
Plato and Mimetic Evanescence 165

sonic structure of the second line with the words préfère and l’Impair
operates as its double echo. Thus the verbal arrangement of the
stanza almost mimics the way air transports sound: one is reached
by the echo before one reaches the source that creates it. This is just
an indicative instance of the many insights that air and airiness
stirred in poetic endeavours of the period. Given that the poem is
entitled Art Poétique and that the lines quoted above form its first
stanza, one may clearly sense the emphasis put by Verlaine on
atmosphere primarily in its literal but certainly in its metaphorical
meaning as well.
It is tempting to contemplate along the same lines one of the most
interesting vase paintings associated with Sappho and her circle in
antiquity. On a hydria displayed in the National Archaeological
Museum of Athens a seated woman, who is surrounded by three
standing ones, is holding a book roll displaying four legible words:
θεοὶ, ἠερίων ἐπέων ἄρχομαι. The genitive ΣΑΠΠΩΣ inscribed on the
vase has been generally interpreted as designating this seated (and
reading) woman. Thus the vase, dated between 440 and 430 BC, seems
to represent an Athenian fantasy of Sappho about to read and per-
form her own poems in the company of other women.4 The vase
poses many questions that have received a wide range of stimulating
answers but the one that is relevant to our discussion is the phrase
inscribed on the scroll and especially the syntagma ἠερίων ἐπέων.
Ἠερίων is an attempt to render in Ionic dialect what in Sappho’s
Lesbian-Aeolic would have been ἀϝερίων, meaning aerial, ethereal.5
Interestingly, two other words, ἔπεα and πτερόε<ν>τα are written on
the left and right rolled parts of the scroll respectively. It has recently
been suggested that the juxtaposition of the familiar Greek epic
phrase ἔπεα πτερόεντα (winged words) with the unfamiliar ἔπεα
ἠέρια (aerial, ethereal words) is probably a meaningful juxtaposition
on the part of the painter, perhaps to be explained as a contrast
4
On the hydria see Edmonds (1922) 1–14. For a relatively recent extensive
discussion of this vase see Yatromanolakis (2007) 153–64 with further bibliography.
5
On the Ionicization of ἠερίων see Edmonds (1922) 4 who thinks it indicates the
familiarity of the painter with Homeric language. Yet the Ionic form ἠέριος clearly
meaning airy, air-like is encountered in the Hippocratic corpus (Vict. 10). The
semantics of the Homeric ἠέριος is ambiguous, meaning either ‘of the early morning’
or ‘wrapped in mist’, ‘in the air’ for which see LfrgrE s.v. For a further defence of the
meaning of ἠέρια as an alternative of πτερόεντα (thus meaning ‘in the air’) see
Immerwahr (1964) 47 and n. 1.

between male and female poetry or between an epic expression and
a melic one.6
Given the uniqueness of the phrase ἔπεα ἠέρια in the extant Greek
corpus no conclusive argument can be made as to its exact connota-
tions. Yet there is no doubt that although the adjectives πτερόεντα
(winged) and ἠέρια (aerial, ethereal) evoke similar associations, their
semantic priorities are different. The former’s semantic priority is on
the very medium that enables flight, namely the wings or feathers that
render words similar to birds or arrows.7 Thus the phrase is clearly
meant to evoke associations of movement and directionality.8 On the
contrary, the latter’s semantic priority, indeed its exclusive focus, is
on air itself, on the ethereal substance of words, the aerial matter they
are envisioned to consist of.
If so, one should probably comprehend the word ἠέρια primarily in
its most literal semantic range, very much along the lines Hippocrates
describes air in his treatise Airs, Waters, Places. More than in any
other Hippocratic work, air in this treatise comes up time and again
as the often invisible yet quintessential substance that enfolds earthly
matter and affects every aspect of organic life; the changing ratios of
its consistency modulate the ways in which it is perceived, largely—
but not exclusively—by vision. After all, Greek often employs the
same word to refer both to plain air and to its transmutation into mist
or fog, both of which are regularly denoted by Hippocrates by the
word ἠήρ.9
This is probably the semantic range in which the word ἠέρια
(aerial, ethereal) is to be understood in the phrase ἔπεα ἠέρια in the
aforementioned Athenian depiction of an imaginary Sappho and her
circle on a fifth-century hydria. Especially in a culture where poetry is
performed orally and perceived aurally, words may indeed be sensed
as having the transparency of atmosphere and the lightness of air, by
way of which they reach the ears of the listeners. But there is more to
the phrase ἔπεα ἠέρια for, despite their intangibility, words, like air,
6
In discussing the interesting juxtaposition of ἠέρια with πτερόεντα
Yatromanolakis (2007) 162 is inclined to interpret the former as ‘lofty’. Though this
interpretation cannot be excluded, I think the juxtaposition (and perhaps opposition)
of the two phrases in terms of poetic genre is more likely to put a semantic emphasis
primarily on the airy quality of the melic ἔπεα and not on their loftiness.
7
On these associations see for instance Kirk (1985) 44; Latacz (1968) 27–32.
8
On the directionality of the Homeric phrase see Martin (1989) 26–37.
9
See for instance Hp. Aër. 6.10; 8.30–3.

may be felt—momentarily at least—as affecting one’s whole biology.
Above all, like air, the verbal fabric of poetry may at times be sensed
like a diaphanous layer surrounding, indeed enfolding, the listener.
Although one would hope that the line θεοί, ἠερίων ἐπέων ἄρχομαι
may represent the beginning of a poem composed by Sappho and lost
to us, for the particular set of questions tackled in this chapter its
authenticity does not affect its importance. In other words, even if the
line was not Sappho’s but instead captured the way Sappho’s poetry
was perceived in Athenian elite circles, the issue at stake remains the
same. Was there a particular aspect in the composition and perform-
ance of such melic poetry that would make its verbal texture, its ἔπεα,
be sensed the way air and atmosphere are?
Perhaps the ethereality that poetry’s peculiar music can create
along with the responsiveness that such poetry elicits are recurrent
ideas emerging under differing cultural circumstances and agendas.
Such issues bring us closer to the way in which relatively recent
debates about the role of aesthetics have revived and expanded the
concept of Stimmung. Usually translated into English as mood or
atmosphere, Stimmung became central in musical and more broadly
aesthetic ideologies of the nineteenth century, reaching its peak at the
turn of the century.10 Originating from realizations about music’s
peculiar affectivity and the listeners’ responsiveness to it, it has also
been understood (and translated) as tuning or attunement and asso-
ciated with Romanticism. In the twentieth century its Romantic
origins receded while the concept reemerged quite forcefully in philo-
sophical debates, especially in the work of Martin Heidegger.11
This is not the place to discuss the various debates over the
twentieth century about whether or not and how to employ the
concept of Stimmung or the subtle nuances that have been claimed
to differentiate this concept from the concepts of mood or atmos-
phere. It is important to stress, however, that especially over the last
two decades, there have been vigorous attempts to widen the territory
of aesthetics by exploring the broader field where all three notions
(Stimmung, mood, and atmosphere) belong, namely that of sense
modalities and perception. Although ‘atmospheres are indeterminate
as regards their ontological status’ as Böhme put it, and although ‘we
are not sure whether we should attribute them to the objects or
10
Wallrup (2015) 1–12.
11
Wallrup (2015), esp. 15–68 and 69–109.

environments from which they proceed or to the subjects which
experience them’, ‘they seem to fill the space with a certain tone of
feeling, like a haze’.12
The metaphor of haze takes us back to the various metamorphoses
of air, to be contemplated primarily as a natural phenomenon before
it is regarded as a concept befitting less concrete yet all-pervasive
modes of aesthetic experience. In his relatively recent work on Stim-
mung, Gumbrecht stresses the key role that acoustic stimuli play in the
creation of atmospheres while also asserting that hearing ‘involves the
entire body’ and that ‘every tone we perceive is a form of physical reality
that “happens” to our body and, at the same time, “surrounds” it’.13
More importantly, Gumbrecht brings up the interrelation, indeed the
deeper affinities, between experiencing a certain soundscape and experi-
encing a certain state of atmosphere, namely weather. ‘Being affected by
sound or weather, while among the easiest and least obtrusive forms of
exprerience, is, physically, a concrete encounter (in the literal sense of
en-countering: meeting up) with our physical environment.’14
This is an interesting turn in recent attempts to revive and expand
theories of Stimmung while also maintaining and elaborating on
what has always been its core, namely music and musicality. When
it comes to verbal compositions, there is a clear focus on the way
prosodies can create atmospheres and modulate moods.15 Prosodic
textures have of course always been essential in all verbal art, yet, as
noted earlier, there are moments in the histories of cultures where,
under differing agendas and practices, the concreteness of words is
made to (or thought to) evaporate, yielding to purer forms of sensory
perception. In recent history, the lyric poetry that was generated
within the so-called Symbolist movement, with Verlaine’s Art Poé-
tique as emblematic, represents a typical moment of such endeavours.
As has been suggested, ‘to understand the atmospheric relationship of
the ideas and to understand that atmosphere is more important than
the simple explication of the theme is to comprehend Verlaine and the
basic aesthetic principle of Symbolism’.16 Interestingly, Greek lyric
poetry and its performance seem to have posed relevant questions
and, as it turns out, Platonic thought presents a most intriguing
response to the challenges such questions triggered in antiquity.
12 13
Böhme (1993) 114, italics mine. Gumbrecht (2012) 4.
14 15
Gumbrecht (2012) 4. Gumbrecht (2012), esp. 5 and 13.
16
Hertz (1987) 105.

WORDS, MELODY, RHYTHM
Although composers such as Claude Débussy were keen on setting

contemporary lyric poems to music, Symbolism’s preoccupation with
the role of music in poetry focused predominantly on the musicality
and harmony emanating from the phonetic handling of language
itself.17 In archaic and classical times, however, lyric poetry, with
the various melic genres at its core, was sung and accompanied by
actual music. This well-attested and well-known fact should inform
the ways in which we may look for the tensions occasionally erupting
between an emphasis on the semantics of words, on the one hand,
and the sensory thrust, on the other, generated by the sonic fabric of
language and greatly amplified through actual music. The relation-
ship between words and music in melic poetry and, more specifically,
the role of meaning in melic compositions are central issues in the
Republic and, as we will see, come up in other works as well.
In the third book of the Republic, after the much-discussed section
on lexis comes full circle, a new thematic section is introduced in
398c. This new section is dedicated to what Socrates calls ᾠδῆς τρόπος
καὶ μελῶν, in other words the manner of song and μέλη, namely lyric
poetry and performance.18 Soon we are given a very clear definition of
μέλος:
πάντως δήπου, ἦν δ’ ἐγώ, πρῶτον μὲν τόδε ἱκανῶς ἔχεις λέγειν, ὅτι τὸ
μέλος ἐκ τριῶν ἐστιν συγκείμενον, λόγου τε καὶ ἁρμονίας καὶ ῥυθμοῦ.
‘At all events’ I said, ‘presumably, to start with, you have enough of an
understanding to say that lyric verse consists of three elements: words,
melody, rhythm?’ (tr. Chris Emlyn-Jones and William Preddy)
Interestingly, in the course of the discussion that takes place between
this introductory definition of μέλος and the marked end of the
section in 403c we hear no fewer than three times that words—for
which the term used is λόγος—should always play the primary role in
the case of μέλος while ἁρμονία and ῥυθμός should just follow it. The
first such statement comes immediately after Socrates has defined
μέλος (398d), the discussion at that point focusing on the verbal
17
On Débussy’s interest in setting poetry into music in relation to the symbolist
aesthetic see for instance Hertz (1987) 85–133.
18
On this section of the Republic see for instance Barker (1984) 128–38;
Moutsopoulos (1989) 67–80; Pelosi (2010) 32–67.

aspect of laments and dirges, of which the corresponding ἁρμονίαι
have been banned earlier in Book 3. The second statement is made in
the context of the broader discussion about the appropriate rhythms
to be practised in musical education—here elaborate rhythms and
movements are rejected while those of ‘an orderly and manly life’ are
recommended. ‘In view of this’, we are told, ‘the metrical foot and
melody must follow the verbal expression (λόγος) and not the λόγος to
follow the metre and melody’ (400a). Finally, in the same book and
thematic section, still as part of the discussion about the foundational
role of μέλος in the ideal musical training, we are reminded once
again of the priority of λόγος and its decisive role in the shaping of the
proper μέλος (400d):
And another aspect of what is good rhythm and bad rhythm and what
isn’t: the first resembles and matches fine language [λέξις], the other
does the opposite, and the same applies to what is melodious and what
isn’t, if rhythm and melody match the words [λόγος], as was said just
now, and not the other way round.’
‘Yes indeed’ he said ‘these must match the words [λόγος].’
(tr. Chris Emlyn-Jones and William Preddy)
The primacy or, more accurately, the supremacy of the verbal com-
ponent in melic compositions over the other two, rhythm and har-
mony, appears to be a consistent, often implicit, concern in several
Platonic dialogues, but its emphatic appearance three times in this
section of the Republic within a few pages is remarkable. The
emphasis might indicate, of course, Plato’s hostility towards current
Athenian musical practices, some of which were probably associated
with the legacy of the so-called new musical experimentations that
were prevalent in the last part of the fifth century. It does seem indeed
that, among other bold undertakings, such experimentations would
appear to favour rhythm and harmony over and above λόγος.19 Yet
apart from the consequences of the ‘new musical’ legacy, I think that
Plato’s anxiety was additionally caused by a tendency that was nat-
urally ingrained in the experience of μέλος in general. I am referring
to the inherent tendency of the purely musical components of μέλος,
namely rhythm and harmony, to overpower its verbal component.
Quite a few instances in the Platonic corpus reveal Plato’s aware-
ness of this deep-rooted prevalence of harmony and rhythm. The
19
On this issue see n. 2.

following one comes from the same section of the third book of the
Republic, soon after Socrates’ last warning regarding the necessity for
the supremacy of λόγος (401d–402a).
ἆρ’ οὖν, ἦν δ’ ἐγώ, ὦ Γλαύκων, τούτων ἕνεκα κυριωτάτη ἐν μουσικῇ

τροφή, ὅτι μάλιστα καταδύεται εἰς τὸ ἐντὸς τῆς ψυχῆς ὅ τε ῥυθμὸς καὶ
ἁρμονία, καὶ ἐρρωμενέστατα ἅπτεται αὐτῆς φέροντα τὴν εὐσχημοσύνην,
καὶ ποιεῖ εὐσχήμονα, ἐάν τις ὀρθῶς τραφῇ, εἰ δὲ μή, τοὐναντίον; καὶ ὅτι αὖ
τῶν παραλειπομένων καὶ μὴ καλῶς δημιουργηθέντων ἢ μὴ καλῶς φύντων
ὀξύτατ’ ἂν αἰσθάνοιτο ὁ ἐκεῖ τραφεὶς ὡς ἔδει, καὶ ὀρθῶς δὴ δυσχεραίνων
τὰ μὲν καλὰ ἐπαινοῖ καὶ χαίρων καὶ καταδεχόμενος εἰς τὴν ψυχὴν τρέφοιτ’
ἂν ἀπ’ αὐτῶν καὶ γίγνοιτο καλός τε κἀγαθός, τὰ δ’ αἰσχρὰ ψέγοι τ’ ἂν
ὀρθῶς καὶ μισοῖ ἔτι νέος ὤν, πρὶν λόγον δυνατὸς εἶναι λαβεῖν, ἐλθόντος δὲ
τοῦ λόγου ἀσπάζοιτ’ ἂν αὐτὸν γνωρίζων δι’ οἰκειότητα μάλιστα ὁ οὕτω
τραφείς;
‘In that case, Glaucon,’ I said, ‘isn’t the training in μουσική most
sovereign for these reasons, in that rhythm and melody above all plunge
into the innermost soul and take a most vigorous hold of it, bringing
with them the beauty of form; and, if one is trained correctly, they make
him beautiful and good in form; if not, isn’t the result the opposite? And
furthermore he who has been brought up as he should have been, will be
most acutely aware of what has been omitted and not well made, or not
well nurtured, and he would rightly disparage it and approve and rejoice
in what is beautiful, allow it into his soul, feed on it and become a good,
fine man. On the other hand would he rightly reject and hate what is
shameful even while still young, and before he is able to reason these
things out, and, because he has been brought up in this way, when
reason does come would he welcome it because he recognizes its utter
fitness for him?’ (tr. Chris Emlyn-Jones and William Preddy adapted)
The deep immersion of rhythm and harmony into the innermost part
of the soul and their vehement grasping of it, we are told, play a
determining role in one’s formation. At the same time, what is empha-
sized here is not only the physical and mental potency of rhythm and
harmony but also their antecedence in respect to λόγος—understood in
this case not just as language but as fully fledged discursive capacity.
It is precisely this dominant rootedness of rhythm and harmony,
I claim, eloquently and powerfully acknowledged by Plato himself
here, that haunts his thought and turns μέλος into a field of pro-
foundly ideological contestation. To put it differently, underneath
Plato’s repeated warning that in melic genres λόγος ought to have
the absolute priority, with rhythm and harmony second in order, one

senses the fear of a deep-seated order, both natural and cultural, that
may in fact be exactly the opposite. What, then, does the deep-rooted
dominance of harmony and rhythm, of which Plato himself is aware,
really threaten in his rigorous, λόγος-ruled, melic decorum? I will
attempt to give an answer to this question in three movements, going
through three sets of passages from different dialogues, which high-
light different aspects of the same underlying phenomenon. The three
sets of passages come from the Republic, the Laws, and finally, the
Phaedrus.
STIMMUNG AND THE RESIDUE OF WORDS
The first passage comes in the third book of the Republic (411a–b)
and, although it is located in a later section of that third book, where
Socrates largely focuses on gymnastic training, it is indirectly, at least,
linked to the overall discussion about μέλος in that book:
οὐκοῦν ὅταν μέν τις μουσικῇ παρέχῃ καταυλεῖν καὶ καταχεῖν τῆς ψυχῆς
διὰ τῶν ὤτων ὥσπερ διὰ χώνης ἃς νυνδὴ ἡμεῖς ἐλέγομεν τὰς γλυκείας τε
καὶ μαλακὰς καὶ θρηνώδεις ἁρμονίας, καὶ μινυρίζων τε καὶ γεγανωμένος
ὑπὸ τῆς ᾠδῆς διατελῇ τὸν βίον ὅλον, οὗτος τὸ μὲν πρῶτον, εἴ τι θυμοειδὲς
εἶχεν, ὥσπερ σίδηρον ἐμάλαξεν καὶ χρήσιμον ἐξ ἀχρήστου καὶ σκληροῦ
ἐποίησεν· ὅταν δ’ ἐπέχων μὴ ἀνιῇ ἀλλὰ κηλῇ, τὸ δὴ μετὰ τοῦτο ἤδη τήκει
καὶ λείβει, ἕως ἂν ἐκτήξῃ τὸν θυμὸν καὶ ἐκτέμῃ ὥσπερ νεῦρα ἐκ τῆς ψυχῆς
καὶ ποιήσῃ ‘μαλθακὸν αἰχμητήν’.
Then whenever anyone lets music entrance his soul with its piping, and
lets it pour into his soul through his ears, as though through a funnel,
the sweet and soft and mournful ἁρμονίαι that we were discussing just
now, and when he uses up the whole of his life humming, enraptured by
[lit. ‘brightened by’, γεγανωμένος] the song, then to begin with, if he has
anything of the spirited element in him, this man will temper it like
iron, and make useful what was useless and hard. But if he persists in
entrancing it without ceasing, he will eventually dissolve it and melt it
away, till he pours away his spirit, and cuts, as it were, the sinews from
his soul, and makes of it a ‘feeble warrior’.
(tr. A. Barker 1984, slightly modified)
This is a remarkable return to an issue that seemed to have been
resolved once and for all in an earlier section of the third book, when
the so called θρηνώδεις, μαλακαί (soft), and sympotic ἁρμονίαι were

erased from (ἀφαιρετέαι) the ideal education as unworthy (ἄχρηστοι,
398e). There is no question, I think, that the sweet, soft, and mourn-
ing ἁρμονίαι Socrates talks about here are the ἁρμονίαι associated with
melic poetry and that the setting he has in mind is predominantly a
sympotic setting (398c). I have extensively discussed elsewhere the
cultural and performative subtext of this key passage and have shown
that the verb μινυρίζειν (of which we have the participle μινυρίζων in
this passage) is often associated in Greek texts with the hushed,
subdued quality of melic or melic-like singing.20 In the Platonic
scholia and in Byzantine lexicographical sources, the verb is glossed
as mourning or as low-voiced, subdued singing (ἠρέμα ἀείδειν). The
first meaning (mourning) is indeed encountered in earlier texts, but
by the end of the fifth century, and especially in Aristophanes, the
verb is used mainly to denote soft, barely articulated singing, almost
whispering, humming, or murmuring and is associated with μέλος in
particular. As far as the other participle in this passage is concerned,
γεγανωμένος, I have suggested that in addition to the usual interpret-
ation as ‘enraptured’ or ‘entranced’ we have to take into account the
literal meaning of the verb γανόω, which means ‘polishing’ of certain
materials, including metals, and is well captured in the scholia to
Plato’s Republic as λελαμπρυσμένος, meaning brightened, sparkling.21
I had then claimed that, for a modern reader, the combination of the
semantics of being ‘enraptured’ and ‘brightened’ ‘sparkling’ can be
more fully understood through James Joyce’s conceptualization of the
‘luminous, silent stasis of aesthetic pleasure’.22
The present focus on the relationship between λόγος and music in
Platonic views of μέλος, however, helps us shed light on additional
semantic aspects. More specifically, μινυρίζειν represents what
I should like to call the vocal residue of language when language has
been infiltrated by melody. Murmuring is the faint vocal echo of the
words that are now dim reminiscences of their articulated, discursive,
state. This hushed, melodious, quality of the vocal is what remains
from μέλος in such conditions. In such cases, then, the semantics of
the verbal component tends to vanish, as it yields to the melodious,
musical texture of the song.
20
For a detailed analysis and discussion of relevant texts see Peponi (2012) 20–3.
21
Scholia in Platonem (Greene) ad Rep. 411, bis; Hesychius s.v.; Photius s.v.
22
Peponi (2012) 14–17.

Without doubt, Plato himself saw this low-voiced, hushed, vocal-
ization of ᾠδή as both the cause and the result of an essentially
slumberous state of consciousness. Prolonged indulging in this enrap-
tured humming is seen by him as threatening to completely liquefy
and melt the θυμός while cutting off the sinews from the soul. This
imagery of slackening and feebleness in the Platonic text captures
very graphically the sense of a somniferous, self-absorbed mood,
permeating and spreading over one’s whole existence. This brings
us back to the concept of Stimmung, especially as it has been re-
interpreted and enriched over the last decades. The concept can
indeed capture quite effectively the distinctive atmosphere that was
allegedly generated by certain types of melic performances in
antiquity. Though it is not unlikely that this atmosphere had enough
potency to spread all over the public space of a theatre or an odeion, it
is reasonable to think that intimate environments had the capacity to
provide a more proper venue for this type of experience, where meaning
in its discursive form as λόγος yielded to less solid and more indefinite
modes of perception and feeling that tended to act by enfolding and
cocooning the listener. It is likely that what Plato calls soft and
sympotic harmonies and the poetry that went with them had a special
effectiveness in pouring out this type of atmosphere and in creating
this type of ambience. His juxtaposition of the low-voiced murmuring
of the song with its sparkling effect (denoted by γεγανωμένος) is an
effective way to capture this special lyric mood as an aura diffused
inside and around the person who indulges in this state.
SOUND AND THE NON-MIMETIC ABYSS
Prolonged exposure to this type of atmosphere, induced by the

performance of certain melic genres, is thus clearly perceived as
intensifying an all-pervasive state of sensuousness while discouraging
engagement in straightforward cognitive activity. Although clearly
aware of its allure, we can foresee that Plato would consider such a
condition of consciousness unacceptable. As the solidity of λόγος
evaporates, signification, along with the representational—that is
mimetic—capacity of poetry, tends to recede. Despite the well-
known and thoroughly discussed perplexities Plato brought up in
several dialogues concerning the dangers of mimesis, he was indeed

in favour of the representational essence of mimesis, provided that it
was the type of mimesis he considered proper. In his last work, the
Laws, this becomes clearer than ever. The approved μουσική of the
Laws (including choral, namely melic, song-and-dance) is conceived
of as profoundly and utterly mimetic.23 This should not be surprising.
Once musical practices are allowed in the Platonic city, they have to
adapt to the approved sets of significations and meanings. Mimesis,
then, understood here as a system of representation, is the very
vehicle that guarantees them.
Which brings us to the next passage, from the second book of the
Laws (669c–670a). The passage can be read as yet another Platonic
attack against the legacy of the New Music and its experimenta-
tions.24 The Athenian interlocutor mentions the dangers of assigning
unbefitting tunes and gestures to the wrong type of individual or
collective musical agents. He also talks about the unacceptable mix-
ture of different sources of sounds, such as animal and human cries—
all of which can be read as variations on themes already discussed in
the Republic.25 But the most astounding assertion here is the last part
of the passage (669d–e):26
ταῦτά γε γὰρ ὁρῶσι πάντα κυκώμενα, καὶ ἔτι διασπῶσιν οἱ ποιηταὶ
ῥυθμὸν μὲν καὶ σχήματα μέλους χωρίς, λόγους ψιλοὺς εἰς μέτρα τιθέντες,
μέλος δ’ αὖ καὶ ῥυθμὸν ἄνευ ῥημάτων, ψιλῇ κιθαρίσει τε καὶ αὐλήσει
προσχρώμενοι, ἐν οἷς δὴ παγχάλεπον ἄνευ λόγου γιγνόμενον ῥυθμόν τε
καὶ ἁρμονίαν γιγνώσκειν ὅτι τε βούλεται καὶ ὅτῳ ἔοικε τῶν ἀξιολόγων
μιμημάτων·
And in the midst of all this confusion, he will find that the poets also
divorce rhythm and movement from the μέλος by putting tuneless
words into metre, and rob μέλος and rhythm of words [ῥήματα] by
using stringed instruments and pipes on their own. When this is done, it
is extraordinarily difficult to know what the rhythm and harmony
without λόγος are supposed to signify and what worth-while mimetic
enactment they depict. (tr. J. Saunders adapted)
23
On mimesis as an acceptable practice in the Laws see for instance Pl. Lg. 655d–
656a; 668b; 795e–796c; 816a.
24
On this see also Kowalzig (2013b) 185–6 and n. 49.
25
In terms of its hints at the various and varying New Musical practices this
passage is complementary to Rep. 397a–c.
26
Text printed as in Burnet (OCT).

This is the most direct acknowledgment of the positive effects of
mimesis in Plato’s thought. Despite some hesitance in the past
regarding the diction and meaning of the passage, it is very clear
that what may appear as lack of mimetic transparency is thought to
cause cognitive chaos.27 Furthermore, it is clear that the presence of
ῥήματα and λόγος is precisely what provides the assurance, or rather
the comfort, of an uncontested mimetic referent.
In other words, what is described here is a crisis very particular to
melic genres. For, clearly, it is only in the case of μέλος that rhythm
and harmony can act in this self-determining, one may say emanci-
pated, manner. Contrary to epic poetry, dramatic dialogues, and
longer elegiac narratives, all of which are evidently tied to the verbal
and thus make immediately apparent what may be wrong about them
from a Platonic perspective, in the case of μέλος the prevalence and,
potentially, the autonomy of the purely musical components may
cause other types of cultural destabilization. For, by promoting dif-
fused atmospheres and moods rather than concrete sets of ideas or
even emotions, they are hard to pinpoint, let alone to manipulate.
If our analysis is legitimate, we can now appreciate a remarkable
formulation, located earlier in the same section of the Laws (669b):
μὴ τοίνυν ἀπείπωμεν λέγοντες τὸ περὶ τὴν μουσικὴν ᾗ χαλεπόν· ἐπειδὴ
γὰρ ὑμνεῖται περὶ αὐτὴν διαφερόντως ἢ τὰς ἄλλας εἰκόνας, εὐλαβείας δὴ
δεῖται πλείστης πασῶν εἰκόνων.
Let us not hesitate, then, to mention the point wherein lies the difficulty
of music. Just because it is more talked about than any other form of
representation (εἰκών), it needs more caution than any (εἰκών).
(tr. R. G. Bury)
In this passage εἰκών is usually translated as representation.28 This is a
fairly satisfactory translation yet it misses the semantic priority that
the term εἰκών puts on visual representation in particular. In other
words, Plato’s choice of the term εἰκών here is deliberately marked. Its
emphatic employment twice builds upon the Athenian’s effort to
capitalize on the idea that μουσική belongs in a system of τέχναι
εἰκαστικαί (667d). This is a remarkable twist in Platonic theory.
27
On textual and interpretive issues see England (1921) 325–7. For other aspects
of this section of the Laws and its relation to broader concerns about musical ēthos see
more recently the interesting discussion by Pelosi (2010) 59–67.
28
See also Saunders (1970) 109, who translates it as ‘artistic representation’. Yet
Pangle (1980) 51 stays closer to the original by translating it as ‘image’.

Although it is well known that Plato (and for that matter Aristotle
as well) was striving to point out affinities between poetic and visual
arts, in this case we are met with a formulation that goes much
beyond similarities or analogies. Instead, what Plato attempts to
do here is to create a complete equation. In brief, μουσική is not
just like an εἰκαστικὴ τέχνη. Quite surprisingly, it is identified as one
of them (668a):
οὐκοῦν μουσικήν γε πᾶσάν φαμεν εἰκαστικήν τε εἶναι καὶ μιμητικήν;
We assert, do we not, that all music is representative and imitative?
(tr. R. G. Bury)
Three points are crucial here. First, the term εἰκαστικός appears very
rarely in Plato’s works—in fact it appears only seven times in the
entire Platonic corpus. Second, it only appears in the Sophist (five
times) and in this part of the Laws (twice). Third, in the complicated
argument of the Sophist the terms εἰκαστικός and εἰκαστικὴ τέχνη
designate visual arts in particular—specific issues of perspective and
symmetry are brought up along with explicit references to painting
and sculpture.29
What prompted, then, this singular emphasis not simply on the
representational but more specifically on the eikastic, visual, quality of
μέλος in the Laws? My answer is that it is precisely Plato’s anxiety
about the non-representational potential of μέλος, its capacity to
provide pleasure beyond fixed systems of signification, that haunts
his theoretical apparatus in this part of the Laws. To put it in a
different way, precisely because the lack of mimetic referent is under-
stood by Plato to be a socially and culturally destabilizing condition,
he insists here on describing μέλος as the most evident and intelligible
case of mimesis, not just as analogous to the visual arts but, quite
extraordinarily, as one of them.
Despite the fact that New Musical experimentations—which, we
should not forget, used a variety of melic forms as their vehicle—increased
29
Pl. Sph. 235e–236a. On the term εἰκαστικός as well as the differentiation between
eikastic/phantastic mimesis in the Sophist see Halliwell (2002), esp. 61–8. It should be
noted that the conceptual differentiation between eikastic and phantastic mimesis in
the Sophist does not affect the point made here. The visual force of the term eikastic is
unquestionable. Furthermore, if we understand the term eikastic in the Laws accord-
ing to its definition in the Sophist (235d–e)—namely as an exact visual reproduction of
an original model—then Plato’s need to turn μουσική into an unambiguous, and thus
cognitively unproblematic, representation comes across even more clearly.

the fear of mimetic and representational destabilization, Plato was
absolutely aware that the non-representational potential of μέλος
along with the dissolution of its semantic constituent were always
lurking, even in more traditional types of performance. As we saw
earlier, his unique description of the listener of the soft, lament-like,
and sympotic ἁρμονίαι in the third book of the Republic, eternally
murmuring the melody while enveloped by its atmosphere, fully
indicates precisely this awareness.
AN ECOLOGY OF MELOS
In order to effectively understand atmosphere as a metaphor, one has

to conceive of it in its most literal, physical, sense. This is why earlier
in this chapter I referred to the Hippocratic Airs, Waters, Places as a
text that fully captures the ways in which air, to be generally under-
stood as atmosphere, envelopes and affects organic matter. I should
now like to suggest that Plato’s Phaedrus is the Platonic work par
excellence where lyric poetry is best conceptualized, or rather sensed,
as atmosphere, while atmosphere in its most literal meaning modu-
lates the dialogue from the beginning to the end. To put it differently,
nowhere else in Plato is the osmotic action between physical atmos-
phere, on the one hand, and atmosphere as mood or Stimmung, on
the other, made so perceptible. And nowhere else in the Platonic
corpus is lyric poetry so consistently and creatively used in order to
emblematize precisely this osmotic action.
The fact that the dialogue is rich in references to lyric poetry is well
known and thoroughly studied, thus unnecessary to repeat.30 The
specific point made here, though, is that understanding the role of
atmosphere as an integral part of aesthetic experience helps us better
grasp not only the reason why lyric poetry is indeed so central in this
particular Platonic dialogue but also the reason why it permeates its
text in so many different ways. More specifically, it has been noted, in
some cases with great insight, that lyric poetry is far more present in
this dialogue than the sum of the explicit references to lyric poets
30
See for instance Ferrari (1990) 86–112; Demos (1999) 65–88; Pender (2007)
1–57; Capra (2014) 27–87.

encountered in it.31 That is, in addition to the passages where lyric
poets are mentioned by name, the text brims with allusions to lyric
compositions. An ancient reader of the Phaedrus familiar with the
poetry performed in sympotic settings would probably sense the
presence of even more lyric allusions spread throughout the text.
Why is μέλος, then, so constitutive in the layering of this dialogue,
both on its surface and in its subtext?
The answer to this question lies precisely in Plato’s general aware-
ness that lyric poetry with μέλος at its core, perhaps over and above
any other genre, tends to create a state of consciousness where the
concreteness of verbal matter may recede, giving way to less palpable
yet wholly pervasive moods or atmospheres that are perceived in
various sense modalities. This is precisely the phenomenon that
makes its appearance in the Phaedrus as well, only here in a predom-
inant and remarkably resourceful way that befits the peculiarities of
the dialogue. Indicative of Plato’s inventiveness and, more particu-
larly, of the way physical and lyric atmosphere are meant to mutually
enhance and redefine one another is the following passage, that
comes in an early stage of the dialogue (235b–e):
ΣΩ. τοῦτο ἐγώ σοι οὐκέτι οἷός τ’ ἔσομαι πιθέσθαι· παλαιοὶ γὰρ καὶ σοφοὶ
ἄνδρες τε καὶ γυναῖκες περὶ αὐτῶν εἰρηκότες καὶ γεγραφότες ἐξελέγξουσί
με, ἐάν σοι χαριζόμενος συγχωρῶ.
ΦΑΙ. τίνες οὗτοι; καὶ ποῦ σὺ βελτίω τούτων ἀκήκοας;
ΣΩ. νῦν μὲν οὕτως οὐκ ἔχω εἰπεῖν· δῆλον δὲ ὅτι τινῶν ἀκήκοα, ἤ που
Σαπφοῦς τῆς καλῆς ἢ Ἀνακρέοντος τοῦ σοφοῦ ἢ καὶ συγγραφέων τινῶν.
πόθεν δὴ τεκμαιρόμενος λέγω; πλῆρές πως, ὦ δαιμόνιε, τὸ στῆθος ἔχων
αἰσθάνομαι παρὰ ταῦτα ἂν ἔχειν εἰπεῖν ἕτερα μὴ χείρω. ὅτι μὲν οὖν παρά
γε ἐμαυτοῦ οὐδὲν αὐτῶν ἐννενόηκα, εὖ οἶδα, συνειδὼς ἐμαυτῷ ἀμαθίαν·
λείπεται δὴ οἶμαι ἐξ ἀλλοτρίων ποθὲν ναμάτων διὰ τῆς ἀκοῆς πεπληρῶσθαί
με δίκην ἀγγείου. ὑπὸ δὲ νωθείας αὖ καὶ αὐτὸ τοῦτο ἐπιλέλησμαι, ὅπως τε
καὶ ὧντινων ἤκουσα.
SO: You go too far: I can’t agree with you about that. If, as a favour to
you, I accept your view, I will stand refuted by all these people—wise men
and women of old—who have spoken or written about this subject.
PH: Who are these people? And where have you heard anything better
than this?
31
See Pender’s analysis in Pender (2007) 1–57.

SO: I can’t tell you offhand, but I am sure I’ve heard better somewhere;
perhaps it was the lovely Sappho or the wise Anacreon or even some writer
of prose. So, what’s my evidence? The fact, my dear friend, that my breast is
full and I feel I can make a different speech, even better than Lysias’. Now
I am well aware that none of these ideas can have come from me—I know
my own ignorance. The only other possibility, I think, is that I was filled
through the ears, like an empty jar, by the streams of others, though I am so
stupid that I’ve forgotten where and from whom I heard them.
tr. A. Nehamas and P. Woodruff (slightly adapted)
I have used the term atmosphere throughout this chapter to capture
the diffusion and evaporation of meaning that may take place in the
experience of lyric. Perhaps Plato preferred the idea of liquefaction
instead of evaporation. For it cannot be coincidental that both times
he refers to μέλος as susceptible to diffusion, in the Republic and in
this famous passage of the Phaedrus, he uses slightly different meta-
phors of liquidity—both times he talks about liquids pouring into
one’s self. In the first case, in the Republic, he employed the metaphor
of the ears as the funnel through which the soft ἁρμονίαι pour into
one’s soul; in the second case, in the Phaedrus, it is the very body of
Socrates, his chest, that is filled like a jar with the streams (νάματα)
coming from elsewhere. But in both metaphors, the one of ethereality
and evaporation, and Plato’s own, the one of liquefaction, it is the
same idea that is captured: the dissolution of the solidity of words as
concrete signifying units.
One may think that the two metaphors represent slightly different
registers of the body’s sensorium. As I mentioned earlier, airiness and
evaporation suggest that one’s whole existence is surrounded and
enfolded by a certain mood elicited by the volatility of words. Plato’s
imagery of liquefaction, on the other hand, suggests that one’s inner
body is filled with their fluid essence. Interestingly, then, while the
compactness of words dissolves, perception is in both cases concep-
tualized in terms of a peculiar, inner or outer tactility. Yet in the
Phaedrus this peculiar moist tactility of others’ words reaching Soc-
rates’ inner body like streams (νάματα) is nothing but an integral,
indeed organic, part of the ecosystem within which the entire dia-
logue takes place. For water fills the place as part of its pronounced
ecology. The two interlocutors are sitting, or rather reclining, at the
banks of the river Ilisus. The words for water (ὕδωρ, ὑδάτιον), river
(ποταμός), stream (νᾶμα) punctuate the text from the beginning to

the end.32 Much before we hear of the streams filling Socrates’ body,
we hear him say how delightful and proper it is for this time of day
and season to wet one’s feet in the water. And, far from accidentally,
νᾶμα, which is used here to describe the dissolved essence of words
reaching Socrates’ inner body, resurfaces at the end of the dialogue as
a description of the place, now called Νυμφῶν νᾶμα.33
Water is absolutely essential to both the literal and the metaphor-
ical atmosphere of the Phaedrus. In the aquatic habitat of the dialogue
the liquefied words of μέλος drench its verbal web. The reader senses
their refreshing presence like the water that strokes the barefoot
Socrates on a summer’s day.
32
Pl. Phdr. 229a; 229b; 230b; 235d; 242a; 242b; 278b.
33
Pl. Phdr. 278b.
Aristotle on Music for Leisure

Pierre Destrée
In Politics 8 Aristotle writes (1338a9–24):

It is evident that we should learn and be taught certain things that
promote leisured activity. And these subjects and studies are under-
taken for their own sake, whereas those relating to work are necessary
and for the sake of things other than themselves. For this reason, our
predecessors assigned music a place in education [ . . . ] music is for
pursuit in leisure (πρὸς τὴν ἐν τῇ σχολῇ διαγωγήν), which is evidently
the very reason our predecessors included it in education. For they give
it a place among the leisured pursuits they considered appropriate for
free people.1
As several interpreters have stressed, what Aristotle calls music ‘for
pursuit in leisure’ (or in ‘cultivated leisure’) is to be taken as the most
valuable ‘kind’ of music (I deliberately put ‘kind’ into inverted com-
mas for the moment).2 The main reason for this is that this ‘kind’ of
music is supposed to be, more than the other kinds, part and parcel of
perfect happiness. But several difficulties remain as to its exact meaning
and status. First, is it different from the three sorts of ἁρμονίαι or μέλη
that Aristotle examines at the end of his treatise: ἠθικὰ μέλη, which aim
at the moral improvement of the youth; πρακτικὰ μέλη, ‘invigorating’
songs, or tunes, for relaxation; and ἐνθουσιαστικὰ μέλη, aiming at
katharsis? Secondly, what does Aristotle mean when he rather
1
I quote C. D. C. Reeve’s translation of the Politics, sometimes slightly modified.
2
See esp. Nightingale (2004) 240–52; see also Depew (1991) 367; Ford (2004). The
ethical reading Carnes Lord has defended in his 1982 book (see esp. 84–5) has been,
I think definitely, refuted by Too (1998) 87–90 and Nightingale (1996) 39–42; but see
Jones (2012) for a very interesting and forceful reassesment of Carnes’ ethical reading.
184 Pierre Destrée

obscurely says that this ‘kind’ of music ‘contributes something to
φρόνησις’ (1339a25–6)? And thirdly, since becoming a good judge
of this ‘kind’ of music seems to be the central aim of the musical
education given to young people, what does this imply as to its
meaning and importance? These are the main questions I deal with
in this paper.
*
Before I review these difficulties and try to give a solution to them,
I will briefly turn to the origin of Aristotle’s educational program. It is
commonly assumed that Aristotle had Plato’s Laws at the back of his
mind when writing these particular books of the Politics; there is a
great number of similar ideas, and he regularly uses almost the same
words and expressions found in the Laws. But few scholars have
recognized that, despite such similarities, Aristotle’s approach to
music is toto caelo different from Plato’s. One may even get the
impression, as I propose one in fact should, that Aristotle either
refers to or quotes almost verbatim certain key passages from the
Laws and the Republic only to draw from them a completely different
conclusion.
In a recent paper, Anastasia-Erasmia Peponi has provocatively
contended that Aristotle ‘re-aestheticized’ music and dance after
Plato ‘de-aestheticized’ them in the Laws.3 For the Plato of the
Laws, indeed, the citizens themselves must be part of χορεῖαι, i.e.
should dance and sing, during the whole course of their lives, even
when they are getting old. For the oldest citizens, he even recom-
mends that wine be generously poured so they can be reinvigorated
and hence capable of taking part in χορεῖαι. For Aristotle, on the
other hand, music should be learned and played by children only and
never by adult citizens, who are expected to enjoy music played by
others. Plato also holds that people enjoy music and dance, but (as
Peponi forcefully observes) this is not an aesthetic pleasure, in that for
him music and dance are not meant to be enjoyed as a spectacle as
was commonly the case (and which is something that Aristotle will
commend). The pleasure they are meant to derive comes from
μίμησις in the active sense of the term, as it is the citizens themselves
3
Peponi (2013b). On Plato’s views on the aesthetics of music in the Laws under
ethical constraints, see also Gülgönen (2011) and Rocconi (2012); and on Plato and
Aristotle’s respective views, see Schoen-Nazzaro (1978). For Plato’s views on lyric
poetry see Peponi (this volume).
Aristotle on Music for Leisure 185

who are urged to be active performers of mimetic music and dance.
There is, however, a key proviso, namely that the sort of music and
dance they perform should be the sort that aims at improving them
morally. For if the right sort of music is played (in the Republic, Plato
says that this is music in the Dorian and Phrygian ἁρμονίαι, 399a),
and this music ‘represents’ moral virtues, especially courage and
temperance, performers are supposed to become virtuous: they
become courageous and temperate, having enacted and thus incorp-
orated such values. Moreover, since performing this music through
singing and dancing is enjoyable, it is to be supposed that performers
will learn to appreciate these virtues, and get pleasure from practising
them in the real world; this pleasure is in turn the best possible
motivation to practise virtues. And since no one is capable of achiev-
ing perfect virtue, all citizens will need to practise such music again
and again throughout their lives.
To be sure, Aristotle also takes music to be an important, even
crucial part of moral education, and with this in mind he follows
Plato in having all children, i.e. all future citizens, learn to play an
instrument. But while Aristotle agrees with Plato that, by learning to play
an instrument such as the lyre, one thereby learns the virtues, he strongly
opposes the idea of having citizens play an instrument throughout
their lives: citizens ‘should engage in performance while they are
young and stop performing when they are older’ (1340b36–8).4 And
furthermore, if on Aristotle’s view one values music, that is, the kind of
music that aims at παιδεία, for its moral utility, it remains something
4
True, at the very end of his treatise, Aristotle seems to admit that the older
citizens should continue at least to sing: ‘And each individual should undertake what
is more possible and more suitable for him. But possibility and suitability are
determined by one’s stage of life. For example, it is not easy for people exhausted by
age to sing harmonies that are strained—nature recommends the relaxed harmonies
at their stage of life. That is why some musical experts rightly criticize Socrates
because he rejected the relaxed harmonies for the purposes of education, not because
they have the power that drink has of producing Bacchic frenzy, but because like drink
they make us weak. So, with an eye to that future stage of life—old age—children
should take up harmonies and melodies of this relaxed sort’ (1342b17–27). But it is the
only place where he seems to recommend this. Curiously enough, when he talks about
communal meals (Pol. 7.1329b22–1330a8), Aristotle does not mention such a singing
together, as Plato would have done. It may be the case that in fact the paragraph I have
just quoted does not represent Aristotle’s own views, but is part and parcel of
Aristotle’s critique of Plato. Note that Susemihl and Newman have strongly suggested
that this last bit of our extant text should be considered an interpolation; see Newman
(1902) 571–2.
186 Pierre Destrée

that is merely useful; a thing only acquires its full value when it is
appreciated for itself. This is the main normative principle that leads to
the distinction Aristotle makes between music for παιδεία—for the
inculcation of moral virtue—and music for διαγωγή. While we value
music for its utility in pursuit of παιδεία, we only value it fully and for
its own sake when it is played for our leisure and not for any ulterior
goal. Music for leisure must be a ‘kind’ of music that is appreciated for
itself, and not—as is the case with music for παιδεία—for the sake of
something else.5
*
After this brief introductory presentation of the main bone of con-
tention between Plato and Aristotle, let us have a closer look at how
Aristotle, as I suggest, uses Plato’s propositions while drawing a very
different, and in fact the opposite, conclusion. One of the clearest
instances is the following passage, which in effect constitutes the
starting point of his educational programme (1337b25–32):
Reading, writing, and drawing are taught because they are useful for life
and have many applications (ὡς χρησίμους πρὸς τὸν βίον οὔσας καὶ
πολυχρήστους), gymnastics is taught because it contributes to courage,
but in the case of music a problem immediately arises. Nowadays, most
people take part in music for the sake of pleasure. But those who
originally included it as a part of education did so, as has often been
said, because nature itself aims not only at the correct use of work
(ἀσχολεῖν ὀρθῶς) but also at the capacity for noble leisured activity
(σχολάζειν καλῶς): this is the starting point for everything else (αὕτη
γὰρ ἀρχὴ πάντων μία).
As Aristotle makes clear at the outset, leisure must be the highest aim
in education, and music is one activity particularly suitable for that
aim. Philosophy will provide that too, and no doubt at a superior level
to music, but there is no mention of philosophy in Book 8, which
focuses on the education of the young. The other traditional activities
to be taught to children are fundamentally directed at utility (with the
exception of drawing to which I will come back in a moment). These
activities are undertaken for the sake of some specific goal. One of
5
We might think that there is no reason why, say, a conductor of an orchestra
should not enjoy music for its own sake. But given Plato’s insistence that one must
perform music, namely by singing and dancing, in order to improve oneself, we can
understand why Aristotle emphasizes that in order to fully appreciate music for itself
we must be spectators of or listeners to music instead of being actual performers.

Aristotle’s recurring examples is gymnastics, which is useful because
it prepares a youth for courage by enabling him to acquire strength
and endurance. Now what should strike us, as readers familiar with
Plato’s Laws (as Aristotle’s readers certainly were), is the following
sentence, which sounds rather curious in its context: ‘Nowadays, most
people take part in music for the sake of pleasure.’ This is in effect
an idea we find repeated again and again by Plato, who condemns
people who think that what makes music valuable is just pleasure (see
especially his radical condemnation of the ‘New Music’ which on his
account aims at pure sensory pleasure).6 Clearly Plato also thinks that
music is pleasurable, but what makes it valuable and worth teaching to
the young is not its pleasurable effects per se, but the way it conveys the
right representations of virtue. So here Aristotle endorses Plato’s harsh
rebuttal of such a hedonistic justification of musical education. But the
main idea he will develop from this premise is rather different, and
even opposed to what Plato proposes.
As I noted earlier, Plato’s view is that the right way to engage in
music, and the very reason why younger people must learn it, is for
moral purposes. As he explicitly says in the Laws, gymnastics is to the
body what music is to the soul: ‘In practice, instruction will fall into
two categories: gymnastics for the body, and music for the sake of a
well-ordered soul (εὐψυχίας χάριν μουσική)’, (795d6–8).7 In both
cases, the aim should be the health of the part each is concerned
with. It can hardly be a coincidence that Aristotle insists here, at the
very beginning of his analysis of music (where in fact he is not yet
talking about different kinds or purposes of music, but about the
value of music in general), on the way one should think about
the value of gymnastics: this amounts to an implicit but strong
condemnation of Plato’s approach to music. As Aristotle forcefully
states (this is something he has argued for in Book 7), leisure is ‘the
6
See esp. Rep. 404e, which summarizes his rejection of the New Music, which
he compares to sophisticated cuisine: ‘There complexity engendered intemperance
(ἀκολασίαν ἡ ποικιλία ἐνέτικτεν), didn’t it, and here it engenders illness, whereas
simplicity in musical training engenders temperance in the soul, and in physical
training health in the body?’ See also Laws 2.657b, and more generally 657e–659c
for the critique of pleasure as the criterion for judging art works. On Plato and New
Music, see most recently D’Angour (2006a) and Csapo (2011).
7
It is true that μουσική for Plato also includes dance and songs, but the main
concern in the whole passage (Laws 7.795d–797a) is music sticto sensu. On gymnastics
and music, see already Rep. 403c–405a.
188 Pierre Destrée

only principle of everything’, i.e. the final aim one must seek in every
decision or activity that one chooses or recommends. And since
Aristotle logically adds that ‘our nature itself aims not only at the
correct use of work but also at the capacity for noble leisured activity’,
education should provide the young, that is to say the future citizens
of the best possible city, with the capacity of using both work time and
leisure time correctly. And indeed, Aristotle will eventually (from
chapter 5 onwards) differentiate the two domains of music that
correspond to these two areas of life: music for παιδεία, and music
for leisure. But as he makes crystal clear right at the beginning of his
inquiry into education (without even mentioning music for παιδεία),
when talking about the benefit of a musical education in general, the
crucial thing to focus on is ‘the right way to spend one’s leisure time’.
So, properly understood, music must be, at least primarily, for the
sake of leisure; more precisely, musical education must primarily be
directed at making people able to enjoy leisure properly.
Plato does not actually use the words σχολή or ἀσχολία in the same
sort of context, but interestingly enough it is an idea that is very much
present in several passages. The following passage from the Laws
reveals Plato’s position, which is emphatically different from what
Aristotle will argue (7.806d–807d):
Now that our citizens are assured of a moderate supply of necessities,
and other people have taken over the skilled work, what will be their
way of life (τρόπος τοῦ βίου)? . . . Now, do such leisured circumstances
leave them no pressing work to do, no genuinely appropriate occupa-
tion (τοῖς δὴ ταύτῃ κεκοσμημένοις ἆρα οὐδὲν λειπόμενόν ἐστιν ἀναγκαῖόν
τε ἔργον καὶ παντάπασι προσῆκον)? Must each of them get plumper and
plumper every day of his life, like a fatted beast? No: we maintain that’s
not the right and proper thing to do . . . So we must insist that there is
something left to do in such a life, and it’s only fair that the task
imposed, far from being a light or trivial one, should be the most
demanding of all. As it is, to dedicate your life to winning a victory at
Delphi or Olympia keeps you far too busy to attend to other tasks; but a
life devoted to the cultivation of every physical perfection and every
moral virtue (the only life worth the name) will keep you at least twice
as busy. (mod. tr. Saunders—his italics.)
Once citizens are supplied with all necessities and do not need to
work, they are left with the question of how to live their leisured life.
The only correct answer, as Aristotle will also say, is that they need to
try to find the right occupations. According to Plato, these suitable
occupations will consist solely in the cultivation of physical perfection

and moral virtue. For Aristotle, such a way of spending one’s leisure
time is not what we would properly understand by the term ‘leisure’.
As several interpreters have pinpointed, Aristotle uses the term σχολή
in two rather different senses: it may refer to a life stocked with all
the bare necessities, as Plato understands it as well; but in its proper
sense, which Aristotle seems to be the first to propose, σχολή refers to
the leisure time in which activities are done for their own sake, and
never for the sake of something else, as virtuous or political activities
obviously are (e.g. courage in war is used for the sake of peace).8 For
that reason, properly speaking, the virtuous or political life should be
named ἀσχολία, ‘lack of leisure’, since those moral or political activ-
ities are part of the citizens’ day-to-day ‘work life’, or ‘business’
(which is the common meaning of ἀσχολία).
Returning to music, one can clearly see why Plato is so keen to
defend the moral functions of music. If the appropriate leisured life of
the citizens of a perfect city should consist in ‘a life devoted to the
cultivation of every physical perfection and every moral virtue’, the
only music Plato can think of for educating the young towards such a
life cannot but be music for moral improvement—in the same way as
gymnastics should be conceived as aiming at physical improvement.
But for Aristotle it is the other way round. If the appropriate leisured
life, according to his view, is a life where activities must be performed
for their own sake, without any practical consequences (such as helping
one to become morally better or, as he will say later, to relax from hard
work), and if music is to count among such activities, then music must
be enjoyed for its own sake, and not—at least not primarily—for any
external goal such as moral improvement, or relaxation.
Before getting to the heart of what Aristotle will eventually call
‘music for leisure’ at Pol. 8.5, and the problems this poses, let me add
one more remark on his implicit rejection of Plato’s moralistic con-
ception of music. As we shall see, Aristotle is well aware that he is
proposing a new conception of σχολή, as well as a new conception of
music for διαγωγή. Nevertheless, he is also keen to add that such a
conception is in fact rooted in ancient views of music that precede
Plato; this is clear from his remark that people ‘originally included
music as a part of education’ in the framework of leisure. And indeed,
a little further down, he returns to this point, warmly praising Homer
for having a similar conception (1338a21–32):
8
See esp. Solmsen (1964); Demont (1993); and Nightingale (1996).
190 Pierre Destrée

What remains, then, is that music is for pursuit in leisure (ἐν τῇ σχολῇ
διαγωγήν), which is evidently the very reason our predecessors included
it in education. For they give it a place among the leisured pursuits they
considered appropriate for free people. Hence Homer’s instruction to
‘call the bard alone to the rich banquet’. And he goes on to mention
certain others who ‘call the bard that he may bring delight to all’. Else-
where, Odysseus says that the best leisured pursuit (ἀρίστην διαγωγήν) is
when men are enjoying good cheer and ‘the banqueters seated in due
order throughout the hall, give ear to the bard’. It is evident, then, that
there is a certain kind of education that sons must be given not because
it is useful or necessary but because it is noble and suitable for a free
person (ἐλευθέριον καὶ καλήν).
Such an emphasis on Homer, and more generally such high praise for
something ‘ancient’, is sufficiently rare in Aristotle to be worthy of
attention. And indeed, the appeal to Homer must have triggered in
his readers an awareness of the whole sequence in which the above
quoted lines occur:
Then Odysseus, of many wiles, answered him, and said: ‘Lord Alcinous,
renowned above all men, verily this is a good thing (τόδε καλόν), to
listen to a minstrel such as this man is, like unto the gods in voice. For
myself I declare that there is no greater fulfilment of delight than when
joy possesses a whole people, and banqueters in the halls listen to a
minstrel as they sit in order due, and by them tables are laden with
bread and meat, and the cup-bearer draws wine from the bowl
and bears it round and pours it into the cups. This seems to my mind
the fairest thing there is (τοῦτό τί μοι κάλλιστον ἐνὶ φρεσὶν εἴδεται
εἶναι).’ (Od. 9.1–11 tr. A. T. Murray)
What Homer says here is integrated into Aristotle’s scheme: the best
or ‘most beautiful’ thing is for free men to spend their leisured life in
activities such as listening to music sung and played by musicians in
the context of social meetings like the symposium. But if it is true
that this audience is also well aware of Plato’s work, I think there
is much more than just an argument from authority here, as if
Aristotle wanted to defend his own conception of σχολή and the
music that suits this by his appeal to Homer. In effect, people well
acquainted with the harsh rebuttal of Homer’s poetry in Republic 2
and 3 cannot but be reminded that in one quite remarkable passage,
Plato refers to the very same passage from Homer. In Plato, how-
ever, only the verses describing the serving of food and wine are
quoted (2.390a–b):

What about making the wisest man (σοφώτατον) say that the best thing
of all (κάλλιστον πάντων), as it seems to him, is when ‘The tables are
well laden with bread and meat, and the wine-bearer draws wine from
the mixing bowl, brings it, and pours it in the cups’? Do you think that
hearing things like that is suitable for inculcating self-mastery in young
people? (tr. Reeve)
It is remarkable, especially when one contrasts this with Aristotle’s own
use of the same passage, that Plato only focuses on the food and wine,
and does not even mention music. Notable also is the rather ironic way
he presents the ‘the wisest man’: Odysseus, the hero who in fact teaches
us how to become intemperate! Needless to say, Aristotle would have
said that Odysseus was truly wise in considering that listening to music
is the appropriate activity for free and truly happy men.
*
So far, Aristotle has only presented music in a general way, and has
stressed the idea that music should be an activity which is part of a
leisured life. Now what about the musical education that is supposed
to be given to the young, future citizen of his ideal city? This is the
question Aristotle will review in chapters 5 and 6 (after reviewing
gymnastics in chapter 4). And first of all, he wants to stress on that
music has various aims: it can be enjoyed for the sake of relaxation, as
he already said, for the sake of παιδεία or moral education of children,
and for the sake of leisure (1339a11–26):
As for music, we have mentioned some of the problems in our earlier
discussion, but it will be well to take them up again now and develop
them further, in order to provide a sort of prelude to the arguments that
might be made in an exposition of the subject. For it is not easy to
determine what the power of music is, or why one should take part in
it. Is it for the sake of amusement and relaxation (παιδιᾶς ἕνεκα καὶ
ἀναπαύσεως), like sleep and drink?[ . . . ]Or should we believe instead that
music contributes something to virtue (ἢ μᾶλλον οἰητέον πρὸς ἀρετήν τι
τείνειν τὴν μουσικήν), on the grounds that, just as gymnastics gives us a
body of a certain quality, so music has the power to give us a character of
a certain quality, by instilling the habits that enable us to enjoy ourselves
in the right way? Or does music contribute something to leisure and to
intelligence, for indeed this must be set down third in addition to the
possibilities we have talked about (ἢ πρὸς διαγωγήν τι συμβάλλεται καὶ
πρὸς φρόνησιν—καὶ γὰρ τοῦτο τρίτον θετέον τῶν εἰρημένων)?
On first reading, Aristotle seems to present music for διαγωγή en
passant as a third possibility, while he seems to lay emphasis on music
192 Pierre Destrée

for παιδεία. And indeed he will later insist, as he proposes here, on the
idea of music’s being helpful in enabling people to learn to enjoy
virtuous emotions (1340a18–28):
But rhythms and melodies contain the greatest likenesses of the true
natures of anger, gentleness, courage, temperance, and their opposites,
and of all the other components of character as well. The facts make this
clear. For when we listen to such representations our souls are changed.
But getting into the habit of being pained or pleased by likenesses is
close to being in the same condition where the real things are con-
cerned. For example, if someone enjoys looking at an image of some-
thing for no other reason than because of its shape or form, he is bound
to enjoy looking at the very thing whose image he is looking at.
This sounds exactly like what Plato says in both the Republic and the
Laws. Since music is naturally enjoyable, and provided it consists in a
mimesis of virtuous emotions, one may suppose that the recipient will
enjoy being emotionally involved in this mimesis (whatever this
might actually consist in) and that he will enjoy acting virtuously in
exactly the same way one enjoys seeing someone having previously
enjoyed looking at a portrait of that person. But as I noted above,
Aristotle does not hesitate to compare this to gymnastics. This kind of
music must forge the ἦθος of the soul in the same way as gymnastics
are supposed to mould the qualities of a body. Given what Aristotle
has just said in his general presentation of music, it can hardly count
as something laudatory; although it is something required for becom-
ing a morally good citizen, the aim of this music remains purely
practical. There must be another sort of music that corresponds to
another ‘benefit’ which would not be practical in the way gymnastics
and music for παιδεία are: Aristotle makes this clear when he points
out that ‘we must set out a third possibility in addition to the ones we
have been discussing’. By this, Aristotle firmly states the conclusion of
the central argument he has been defending all along: leisure activities
are the best human activities that provide perfect happiness, and since
music must be part of happiness, there must be a ‘kind’ of music that
perfectly suits leisure time.
Before turning to what exactly this ‘kind’ of music is supposed to
consist in, let us first consider the question of why the young should
learn to play an instrument. Why shouldn’t learning to listen to music
be enough? After all, music gives each person a natural pleasure; slaves
and even some animals can enjoy music (1340a2–3; 1341a15–17). So it

does not seem necessary to be a musician oneself in order to be able to
enjoy music. Aristotle himself raises this central question, beginning
with the problem of music for παιδεία (1339a42–b4):
Why should they learn it themselves, rather than being like the Spartans,
who enjoy the music of others in the right way and are able to judge it?
For the Spartans do not learn it themselves, but are still able, so they say,
to judge which melodies are decent (χρηστά) and which are not.
The main consideration at issue here is how to become a judge of

music. This is the aim of musical education, as Aristotle will forcefully
repeat again and again throughout this text. Like Plato, the Spartans
rightly thought music should play an important role in the inculca-
tion of virtue; but unlike Plato, they thought that one does not need to
learn to play an instrument oneself in order to do so, and to become
able to judge between morally good and bad music. This, Aristotle
repeats, is simply not the case: ‘It is difficult if not impossible for
people to become excellent judges of performance, if they do not take
part in it’ (1340b23–5).
But what does this judgement consist in? At this level, there is no
aesthetic judgement, but simply the ability to discern which music is
morally good or bad, that is, which music is an imitation of good or
bad qualities. And since music is supposed to be an imitation of good
or bad emotions, we need to be educated in music in the same way
that we learn how to become virtuous. It might be thought that just
learning how to make such distinctions by learning to listen to music
would be sufficient. But Aristotle is convinced that just as one must
learn how to become virtuous by practising virtuous actions, so too
the active practice of music (namely by learning to play an instru-
ment) is the best and easiest way to become a good judge of music.
This entails becoming a good judge as to what sort of music should
count as suitable music for παιδεία.
This conception, to be sure, exactly reflects Plato’s position as seen
in the Laws (2.668c + 2.669a–b):
So it looks as if a man who is not to go wrong about a given composition
must appreciate what it is, because failure to understand its nature—
what it is trying to do and what in fact it is a representation of—will
mean that he gets virtually no conception of whether the author has
achieved his aim correctly or not . . . So anyone who is going to be a
sensible judge of any representation—in painting and music and every
other field—should be able to assess three points: he must know, first,
194 Pierre Destrée

what has been represented; second, how correctly it has been copied:
and then, third, the moral value of this or that representation produced
by language, tunes and rhythms. (tr. Saunders)
Now if what I have argued for so far is on the right track, becoming
such a good judge in the realm of music for παιδεία cannot be the
final aim of musical education; rather, that process must be directed
at becoming a good judge in the realm of music for leisure. I return
now to the passage I quoted from 8.5, where Aristotle asked why the
young should learn to play an instrument, and expands the terms of
his discussion (1339b4–10):
The same argument also applies if music is to be used to promote well-
being and the leisured pursuits appropriate to someone who is free
(πρὸς εὐημερίαν καὶ διαγωγὴν ἐλευθέριον). Why should they learn it
themselves rather than benefiting from the fact that others practice it?
In this regard, we may consider the conception we have of the gods; for
Zeus himself does not sing or accompany poets on the lyre. On the
contrary, we even say that musicians are vulgar craftsmen, and that a
true man would not perform music unless he were drunk or amusing
himself.
Why should one learn to play an instrument in the case of music for
leisure? Why should we not be able to enjoy others playing music
without ourselves having to know how to play? If it is true that Zeus,
our paradigm of happiness, is supremely happy and that he enjoys
music without playing an instrument himself, why should our young
citizens learn to play? And is it not the case that we should consider it
unfitting and disreputable that free citizens should play musical
instruments? Aristotle does not pursue this issue here, but explicitly
postpones the problem, before returning to it at 8.6 (1340b20–5):
We must now discuss the problem we mentioned earlier of whether or
not the young ought to learn to sing and play an instrument themselves.
It is not difficult to see, of course, that if someone takes part in
performance himself, it makes a great difference in the development
of certain qualities, since it is difficult if not impossible for people to
become excellent judges of works being performed, if they do not take
part in it (μὴ κοινωνήσαντας τῶν ἔργων κριτὰς γενέσθαι σπουδαίους).
On a first reading, it might seem that Aristotle is merely repeating his
previous argument: as is the case with music for παιδεία, one must
learn to play an instrument if one wants to become of good judge of
music. But it seems to me that what Aristotle now has in mind is

totally different from what he said before in the context of music for
παιδεία. In this scenario, being a good judge obviously does not mean
being able to discriminate between the morally good kind of music
and the morally bad kind, which is how Plato conceives the process of
judging music. Aristotle explicitly talks about ἔργα to designate the
musical pieces as they are being performed. A little further on, he
adds the following (1340b35–9):
[S]ince one should take part in performance in order to judge, for this
reason they should engage in performance while they are young and
stop performing when they are older, but be able to judge which
melodies are beautiful and enjoy them in the right way (τὰ καλὰ κρίνειν
καὶ χαίρειν ὀρθῶς), because of what they learned while they were young.
Again, it is tempting to read this from a Platonic perspective, espe-

cially given that τὰ καλὰ κρίνειν is the expression Plato uses to
designate the art of judging which music expresses virtues, or ‘noble
actions’, and which does not. But reading the passage in that way
would contradict what Aristotle has defended so far. For he has
claimed forcefully that, when they have become adults, citizens
must devote their leisure time to enjoying music for its own sake; if
τὰ καλὰ κρίνειν were to amount to discriminating between which
songs are morally good and which are not, this would totally under-
mine the ‘for their own sake’ criterion. The expression here should be
taken as conveying a meaning much closer to what modern critics, at
least since Hume, would call something like a judgement of taste,
namely a judgement about how beautiful such and such a piece of
music is, or how artistically well performed a specific musical per-
formance is.
*
Now what exactly is to be judged? Or more precisely, which qualities
or which properties are these well-formed judges of music supposed
to focus on? But first of all, to return finally to our initial question,
what ‘sort’ of music shall we suppose Aristotle is talking about when
he refers to ‘music for leisure’?
Crucial to these questions is a feature of the argument to which
interpreters have not paid sufficient attention. At the end of his treatise,
when Aristotle reviews the kinds of songs, or tunes, that have been
studied and classified by ‘philosophers of music’ (i.e. people who are
professional critics of musical styles and performances), he does not
196 Pierre Destrée

name any special song that would correspond to such ‘music for
leisure’. Indeed, these songs, or tunes, should correspond to the two
other kinds of music he mentioned earlier, plus katharsis (which he just
named a few lines below, at 1341a21–4): the ‘moral songs’ (ἠθικὰ μέλη)
are those that should be used in educational music, the ‘invigorating
songs’ (πρακτικὰ μέλη)9 for relaxation music, and the ‘inspirational’, or
‘ecstatic songs’ (ἐνθουσιαστικὰ μέλη) for music that aims at some
emotional katharsis. Since he seems to endorse this classification, and
he does not name or propose any other sort of melody, it follows
that the melody proper to the music for leisure must be one of these
three. But which one?
One might be tempted to consider that ‘invigorating melodies’
would be a good candidate. As Aristotle seems to be saying when
he presents these songs, or tunes, the aims of music that correspond
to these are moral education (παιδεία), purgation or purification of
emotions (katharsis), and—I quote Andrew Barker’s translation—
‘thirdly . . . amusement for the sake of relaxation and relief from
tension’ (τρίτον δὲ πρὸς διαγωγὴν πρὸς ἄνεσίν τε καὶ πρὸς τὴν τῆς
συντονίας ἀνάπαυσιν, 1341b36–41), apparently considering διαγωγή
to form one aim with ἄνεσις and ἀνάπαυσις. But as several commen-
tators have noticed, not only is the Greek of the last part of the
sentence odd (one would expect something between πρὸς διαγω-
γήν and πρὸς ἄνεσιν), but also this reading would completely under-
mine the very firm and clear distinction Aristotle repeatedly made
earlier between διαγωγή and relaxation.10 Indeed, διαγωγή, as he
previously stated (1339a14–26; 1339b15–19), is not to be confused
with relaxation and its means, παιδία, i.e. amusement (which is, very
confusingly, the term that is used by Barker to render both διαγωγή
and παιδία). Whatever that rather awkward last bit of the sentence
may amount to (I have in fact suggested that πρὸς διαγωγήν should
be best considered a gloss),11 it seems quite unreasonable, given
Aristotle’s insistence on that distinction, to conflate them in one
rather blurred ensemble, and conclude that those invigorating songs
should be part of music for leisure. Additionally, Aristotle said that
9
This is often translated ‘for action’ vel sim., but these songs, according to what
Aristotle says here, should certainly not be employed for any moral ends. The term
must refer to ‘movement’ and ‘action’ in a very broad sense, and perhaps to dance too,
not to action in the specific Aristotelian sense of ‘moral action’.
10
On this passage and its difficulties, see esp. Schütrumpf ’s detailed note
(2005: ad loc.).
11
See Destrée (2017).

this music is for relaxation after hard work, and when he reviews the
‘invigorating songs’ that correspond to music for relaxation, he
repeats that these songs are adressed to βάναυσοι, i.e. proletarian,
working people—so clearly not ‘free’ citizens who do not work and
spend their life enjoying leisured activities.12 True, Aristotle writes
that ‘since theatre audiences are of two kinds, one free and generally
educated, the other boorish and composed of vulgar craftsmen, hired
laborers, and other people of that sort, the latter too must be provided
with competitions and spectacles for the purposes of relaxation’
(1342a18–22), which could imply that free and educated people
may also be offered such kind of spectacle. Admittedly, from the
way the sentence is written one cannot reliably infer whether Aristotle
means this, or rather the fact those free people also need spectacles for
some other purpose, such as leisure. For the sake of consistency,
I think the latter is more plausible; but even if Aristotle meant the
former, namely that even ‘free’ people also need spectacles for their
relaxation, that would not imply that this sort of music should be
part of spectacles where leisure, not relaxation, is at stake.
As to ‘inspirational’, or ‘ecstatic’ songs, one may also be tempted to
take them as a possible candidate for music for leisure. For Aristotle
insists that everyone, not only deeply unbalanced people (τοὺς ὅλως
παθητικούς), can enjoy them (1342a11–14). But saying that everyone,
including ‘free’ citizens may from time to time be in need of some sort
of ‘curative’ music concert to regain their emotional balance, does not
mean that katharsis should be part of leisure: spectacles that are
supposed to end up with some sort of emotional purification are
simply another venue for enjoying music. Moreover, and more cru-
cially, as Aristotle had stressed ealier, this cathartic (or as he there
says, ‘orgiastic’) music is to be played on an aulos (1341a21–4), while
the musical instrument the citizens are supposed to learn to play in
their youth is the lyre, not the highly emotive aulos which is not
suitable for decent citizens. Since Aristotle insisted that one must
learn to play an instrument in order to become a good judge in music
for leisure, and that that instrument should be the lyre and not the
aulos, it would be very strange if he now considered music for aulos to
be the sort of music one should listen to for leisure.
12
I read with Sauppe (followed also by Ross and Barker) πρακτικά at 1342a15
instead of MSS καθαρτικά. Following Schütrumpf (see his very helpful note ad loc.),
I take 1342a16–28 to be the explanation of how those ‘reinvigorating songs’ contribute
to relaxation.
198 Pierre Destrée

Thus, the only possible candidate for music for leisure must be
the ‘ethical songs’, or rather, as Aristotle says, the ‘most ethical songs’,
or ‘tunes’ (ἁρμονίαις . . . χρηστέον . . . πρὸς μὲν τὴν παιδείαν ταῖς
ἠθικωτάταις, 1342a1–3), i.e. the Dorian tune which is ‘the steadiest
and has a more courageous character than any other’ (περὶ δὲ τῆς
δωριστὶ πάντες ὁμολογοῦσιν ὡς στασιμωτάτης οὔσης καὶ μάλιστα ἦθος
ἐχούσης ἀνδρεῖον, 1342b12–14), and other such tunes—one may think
of the Hypodorian which the Problemata present as having ‘a mag-
nificent and steadfast character’ (ἡ δὲ ὑποδωριστὶ μεγαλοπρεπὲς καὶ
στάσιμον, 19.922b14–15) and also perhaps the Lydian one which is
supposed to give some ‘order’, or ‘orderliness’ (τὸ δύνασθαι κόσμον τ᾽
ἔχειν ἅμα καὶ παιδείαν, οἷον ἡ λυδιστί, 1342b31–2).13 This conclusion
follows naturally from our earlier observations: since the only music
children must learn to play is ‘ethical’ music, it must also be true that
if they learn this in order to become good judges of music when they
get older, the music they are supposed to enjoy during their leisure
time must be that same sort of music. Now we also noted that
Aristotle opposes Plato in relation to the aim of leisure activities,
which must be undertaken for their own sake, and not for any external
reason such as moral education or improvement. Taking these two
requirements into account, the implication of Aristotle’s argument is as
follows: music for leisure must consist in listening and enjoying the
same kind of music, i.e. ‘ethical songs’, but under another description.
Now that they have become judges of that sort of music, adult citizens
can enjoy it qua music for leisure, that is, for its own sake.14
13
On this, Aristotle resolutely opposes Plato, who considered those tunes to be
effeminate, not worth a decent citizen (see Laches 188d–e; Rep. 398d–e; and for a
defence of Plato’s view on this, see Ps-Plutarch, De mus. 15–17). Relying on Laws
7.802e, where Plato talks about tunes (without naming them) that are supposed to
help men to acquire ‘magnificence and courage’ (τὸ δὴ μεγαλοπρεπὲς οὖν καὶ τὸ πρὸς
τὴν ἀνδρείαν), and those that fit women for their acquiring ‘moderation and restraint’
(πρὸς τὸ κόσμιον καὶ σῶφρον), one may infer that for Aristotle, moderation and
restraint is what these ‘feminine’ tunes should offer to young male citizens too,
perhaps as a sort of balance to magnificence and courage they get from more ‘male’
tunes, such as the Dorian. (On the meaning of κόσμιος, see Pol. 6.1321b7–8, where
πρὸς εὐταξίαν καὶ κόσμον seems to refer to the discipline and orderliness of the city, or
the citizens).
14
Perhaps one might compare this, mutatis mutandis, to the music of composers
such as Bach, or to take some more contemporary composers, Arvo Pärt, Krzysztof
Penderecki, or Olivier Messiaen. If you are a serious, committed Christian, I suppose
you might take this music, as these composers themselves have certainly taken it, as
both capable of forming the young to the Christian faith and religious contemplation,

*
What does this amount to exactly? Aristotle unfortunately does not
say explicitly, but there are a few passages which give us a clue as to
what he may have had in mind or what he may (or should) have
elaborated on. When presenting what should and should not be
taught to the young, Aristotle emphasizes two main points. First, as
we have seen, he argues that the young should not be taught to play
simply any kind of music, but only the kind that suits παιδεία. More
precisely, they should not be taught to play the emotional aulos, nor
the more difficult kinds of instruments which professionals play, such
as the kithara (1341a19), and other instruments (pectis, barbitos, as
well as heptagonon, trigonon, and sambuke, 1341a40–b1). Secondly,
they should do so in order to be able to enjoy music properly, and not
in the way uneducated people, or even slaves or certain animals, enjoy
it (1341a9–17):
This could be achieved where lessons in music are concerned if the
students do not exert themselves to learn either what is needed for
professional competition or the astonishing or out-of-the-ordinary
works which have now made their way into competitions and from
there into education, but learn the ones not of this sort only up to the
point at which they are able to enjoy noble melodies and rhythms, and
not just the common charm of any music, which appeals even to some
of the other animals, and to the majority of slaves and children as well.
We have seen that judging in the case of music for leisure amounts to
appreciating a good performance of music in the same way as we
nowadays talk of connoisseurship in music. Connoisseurs do not only
enjoy the ‘common charm of music’; they also enjoy music for its
intrinsic qualities, or the qualities of the performance. But contrary to
what we moderns might be tempted to say on Aristotle’s behalf, this
cannot apply to each and every sort of music, not even to the more
complicated kinds. Here one must recognize, however reluctantly,
that Aristotle seems to share Plato’s condemnation of the so-called
‘New Music’ that was being played on a number of more technically
demanding instruments which were not the ones young citizens were
supposed to learn to play. And as we may infer from the last quota-
tion, this is because he thinks, like Plato, that such music is only
emotional, and gives a purely sensual pleasure. Aristotle does not talk
and as constituting the right sort of music a Christian adult might enjoy for its
own sake.
200 Pierre Destrée

explicitly about this ‘New Music’, but one may guess that his reasons
for condemning it are similar to those he puts forward when con-
demning the use of the aulos.15 His main reason is that this instru-
ment, however perfectly it may be suited to stirring the emotions in
the case of music for katharsis, and however much pleasure it gives,
must be rejected from the musical curriculum for young citizens
because it does not develop their intelligence. Other, more demand-
ing instruments are like the aulos in this regard, in that they ‘enhance
the pleasure of people who listen to their practitioners’ (1341a40–b1);
in other words, they give sensual or emotional pleasure and nothing
else. He then adds that ‘the story told by the ancients about the aulos
is also plausible. They say that Athena invented the aulos, but dis-
carded it. There is nothing wrong with saying that the goddess did
this out of annoyance at how aulos-playing distorted her face, but the
more likely explanation is that the aulos does nothing to promote
intelligence, whereas we attribute knowledge and art to Athena’
(1341b2–8).
This last detail is of great interest for our problem, especially since
it may sound, at least from a Platonic perspective, rather unexpected.
What one would have expected is that such instruments do not teach
moral virtue as Plato would have said, and as indeed Aristotle told us
a few lines above. Here the reason why the aulos should not be part of
the musical education of the young is that it prevents the exercise of
their intelligence, the term intelligence (διανοία) here being associated
with science (ἐπιστήμη), the activity to which the most gifted citizens
of Aristotle’s ideal city are expected to devote themselves. It also reso-
nates with the only description, however brief, Aristotle gives us of the
‘music of leisure’ that he advocates: this is a music ‘for intelligence’ (πρὸς
φρόνησιν, 1339a25–6). But what does he mean by this phrase?
Several interpreters have understood φρόνησις here as if Aristotle
meant practical wisdom or prudence.16 And since I have argued that
the only kind of melody that could correspond to this music is the
15
An external proof for Aristotle’s traditionalist attitude vis-à-vis New Music
comes of course from his pupil Aristoxenes’ explicit condemnation of it and nostalgic
praise of the ‘ancient music’. See further Power (2012b).
16
Reeve translates it ‘practical wisdom’; Lord, ‘prudence’. See however Kraut, who
rightly states that ‘making music is a way of exercising the virtue of wisdom, and this
role is to be distinguished from the contribution it makes to the ethical virtues’ (1997:
178), where ‘wisdom’ is to be taken as a theoretical virtue.

moral one, this may seem like the safest reading. But this interpret-
ation would totally contradict what Aristotle has argued for, namely
that music composed to be listened to in leisure must not have any
practical consequence or benefit. It is only when enjoying this music
that the citizens of his ideal city can really enjoy their leisure, a type of
activity that is opposed to their work time which includes political
business as well as more general moral concerns. Therefore, as I have
proposed, enjoying music as a leisure activity must entail enjoying
these ‘ethical songs’ from another perspective. Which perspective? As
Aristotle says here, from the perspective of φρόνησις or, according to
the Athena passage, διανοία, that is ‘theoretical’ intelligence. But how
should we interpret this proposition?
One interpretation consists in claiming that enjoying music in
leisure time should be supposed to enhance our intelligence, that is,
help citizens to train themselves intellectually in order to become
philosophers properly speaking.17 But again this would contradict
what Aristotle has said. Listening to this music for the sake of
enhancing the intelligence as a preparatory way toward philosophical
contemplation would prove something useful, that is, for the sake of a
further end, not something that is worthwhile in and of itself. If we
stick to Aristotle’s own logic and his emphasis on leisured activities
that are undertaken for their own sake, the only coherent way of
reading this is to consider such music for leisure to be an activity
through which our ‘theoretical’ intelligence exercises its power.
Aristotle does not elaborate on this theme, as crucial as it must
have been for him. But the following passage may hold a clue as to
how he conceived the question of how music for διαγωγή related to
intelligence (1338a40–b4):
Furthermore, it is clear that children should be taught some useful
subjects (such as reading and writing) not only because of their utility,
but also because many other areas of study become possible through
them. Similarly, they should be taught drawing not in order to avoid
making mistakes in their private purchases or being cheated when
buying or selling products, but rather because it makes them contem-
plate the beauty of bodies (ποιεῖ θεωρητικὸν τοῦ περὶ τὰ σώματα
κάλλους). It is completely inappropriate for magnanimous and free
people to be always asking what use something is.
17
This has been defended by Depew (1991): see esp. 371–4.
202 Pierre Destrée

As Aristotle repeats again and again, education must aim at some
utility, but should not aim at this exclusively. As we have seen, a ‘free’
education given to ‘free’ men must also include subjects that are
learned in order to be able to enjoy certain activities for their own
sake. All this he argues (implicitly) against Plato, who vigorously
defended a musical education which only aims at moral improve-
ment. (We may even see in the last sentence just quoted—‘It is
completely inappropriate for magnanimous and free people to be
always asking what use something is’—a rather harsh, ironic rebuttal
of Plato). But this passage may also help us to grasp what Aristotle
may have hinted at when saying that music for διαγωγή is music for
intelligence. Indeed the reason why children must learn drawing is
because acquiring that skill will help them to contemplate beautiful
bodies properly; they will thus become good judges who will properly
appreciate the beauty of bodies. Since this is to be taken as an analogy
of musical education, one should conclude that musical education
must be taken as helping citizens to enjoy the beauty of music properly.
‘Beauty’, as Aristotle states in his Metaphysics, consists essentially in
order, symmetry, and definiteness (Met. M 3, 1078a36–b1). We can
infer then that the way music for διαγωγή must be enjoyed just consists
in that: the intellectual enjoyment of its order, symmetry, and
definiteness.18
18
I presented an ancestor of this paper at the Oxford meeting that Armand
D’Angour and Tom Phillips organized in July 2014. I am very grateful to them for
their kind invitation, and to them and their audience for their comments. Special
thanks are due to Andrew Barker for his very challenging critiques of some of my
views (to which I hope to have responded in a convincing way), and to David Creese
and the Editors for their insightful suggestions on a penultimate version of this
chapter.
Sounds You Cannot Hear

Cicero and the Tradition of Sublime Criticism
James I. Porter
Is it her singing that enchants us, or is it not rather the solemn

silence that envelops her weak little voice?
F. Kafka, ‘Josefine, die Sängerin oder Das Volk der Mäuse’
I want in this essay to sketch out a convergence of interests that at

first glance may seem unlikely: between Cicero’s treatise Orator and
Greek euphonistic literary criticism, between Platonism and materialism,
and between the sound of the voice and the sound of a voice you cannot
hear. My aim will be to show how these specific contrasts end up flowing
into a theory of the sublime that has significant parallels with the
treatise On the Sublime attributed to Longinus. My triple thesis, which
is both historical and conceptual, runs as follows. It is that the Longinian
sublime is constituted in part by a synthesis of the very contrastive
sources I just named; that both Cicero and the euphonist critics known
as the κριτικοί, who are reported by Philodemus before him, anticipate
central aspects of Longinus’ theory; and finally that Cicero may be
dependent on the κριτικοί for part of his theory but not necessarily for
the convergences just described. That is, the κριτικοί arrive the same
point, at the same supreme aesthetic value as Cicero arrives at in his
Orator, but by way of an opposite route. He takes the high road of
Platonism, they take the low road of linguistic materialism; but they
meet in the same place, in a theory of the sublime, which presses
the limits of language and of aesthetic and rhetorical theory to a
radical extreme.
204 James I. Porter

One upshot of this kind of pressure being applied to limits of
language, aesthetics, and rhetoric is that language is driven into a
realm that resembles music more than words, a condition that in the
cultures of Greece and Rome is regularly expressed by the notion that
prose, which is spoken, comes to resemble poetry, which is sung. The
voice has a ‘musicality’ of its own that is consistently enchanting and
often sublime.1 Another is that silence plays a greater role than can be
imagined, both in the production and reception of sound. On the
receiving end of things, what one does not hear is as important as
what one hears: in such cases imagination overtakes listening. On the
producing end of things, sounds disguise the process of sound-
production: here, art is the art of dissimulating an experience of sound
that has no obvious correlate in the sounds that go into the experience.
These complications of the auditory experience were well known
among critics and rhetoricians, who regularly analysed and exploited
the fact that some of the most compelling listening experiences rest
on inaudible foundations, on sounds you cannot hear. The same can
be said of the musical theorists who paved the way for later theories of
the read and spoken voice: sound made rhythmic is constituted by
intervals that cannot strictly be heard. I believe that these insights are
rooted in the song cultures of antiquity themselves starting with
Homer, for example in his description of the Sirens, whose seductive
songs are never actually heard, let alone reported. In the place of
sound or song, what the Sirens offer is the promise of song. And
therein lies their seductive allure, and their unsurpassed sublimity.2
At most a sketch of the argument will be possible here. I will begin
with Cicero, who sets up the problem, and then I will turn to the
euphonist critics known to us from Philodemus as οἱ κριτικοί. Longinus
will be appearing backstage and in the wings throughout.
1
Gorgias was one of the first to explore this possibility, but he was by no means the
first. See Arist. Rh. 3.1–9. On the history of ‘the music of the voice’ in antiquity, see
Porter (2010) ch. 6.
2
Odyssey 12.184–201. See Salecl (1998) ch. 3, esp. 60–1, for a powerful reading of
the Sirens as a condensed emblem of the sublime voice of Homer’s poetry tout court—
a point that was firmly grasped by Kafka in his extraordinary parable ‘The Silence of
the Sirens’ (reprinted in Steiner and Fagles, eds. (1962) 98–9). Further, Vermeule
(1979) 203; Pucci (1979); (1987) 209–13; Schur (2014). It is noteworthy that a number
of the visual depictions of the Sirens from antiquity show them playing musical
instruments (a lyre or a syrinx) rather than singing.
Sounds You Cannot Hear 205

CICERO’S IDEAL ORATOR
Cicero’s Orator (46 BCE) opens with a quandary that it subsequently

circles around. The request by Marcus Junius Brutus for a fuller
account of the summus and perfectus orator prompts in Cicero the
following reflection: How can we tell what is the best form of oratory
given the variety of good orators in the world? What is it about them
all that makes them good? The quandary conceals a paradox: what
they share exists in no concrete orator; the perfection that they
variously exhibit is incomplete except in an ideal form that appears
nowhere empirically; and it is in virtue of this ideal that we judge
existing orators to be good (that is, as less than perfect).3 Unless
Cicero can provide an account of the ideal form of oratory, the
general standard for judging orators will appear arbitrary. If he
succeeds, how will any aspiring orator not be discouraged from
inevitably falling short? (3).
Let us look more closely at the paradox that lies at the heart of
Orator, as this is announced in section 7: ‘in delineating the perfect
orator I shall be portraying such a one as perhaps has never existed.
Indeed I am not inquiring who was the perfect orator, but what is the
unsurpassable ideal which seldom if ever appears throughout a whole
speech but does shine forth (eluceat) at some times and in some places,
more frequently in some speakers, more rarely perhaps in others. But
I am firmly of the opinion that nothing of any kind is so beautiful as
not to be excelled in beauty by that of which it is a copy, as a mask is a
copy (imago) of a face.’ Cicero’s ideal of the summus orator is Platonic
philosophically speaking but not aesthetically speaking (10: has rerum
formas appelat ἰδέας . . . magister Plato).4 It is an idea that leads the
mind to ever higher heights of aesthetic imagination, not to the
abandonment, à la Plato, of the aesthetic domain in the name of a
3
Quaeris . . . quod eloquentiae genus probem maxime et quale mihi videatur illud,
quo nihil addi possit, quod ego summum et perfectissimum iudicem (Or. 3); cf. 55.
I follow Kroll’s revised text of Jahn (Kroll (1913)) and Hubbel’s Loeb translation.
Similarly, Quint. 1.1.19: ‘Such a person has perhaps never yet existed; but that is no
reason for relaxing our efforts to attain the ideal (ad summa tendendum est)’;
tr. Russell.
4
It is no doubt this odd disjunction that causes Kroll in his commentary (1913) to
berate Cicero for sloppy philosophizing; see Kroll ad Or. 8, where he calls the appeal to
Platonism ‘feuilletonistisch’.
206 James I. Porter

metaphysical one. But Cicero’s ideal is Platonic in a further sense: it
leads the mind to abandon the senses and every sensuous image of
beauty in the very imagination of a more perfect aesthetic beauty.
Confronted with a statue of Phidias or a painting by Apelles, no
matter how unsurpassed these may seem, we can always ‘imagine
(cogitare) something more beautiful’. What we see is what the artist
saw. We view the vision he had. And the content of that vision is an
‘intellectual ideal,’ a cogitata species, or a mental form (9).
This last move neatly saves Cicero from a circularity. Or else it
defers it. Instead of having to explain how it is that we can judge any
one work of art according to a standard that no work of art physically
embodies, he only needs to explain how it happens that artists and
viewers come to have in their minds the same ideal standards of
beauty. For that, we have to consult Plato and the later Academy;5
Cicero takes us no further. But he is not out of the woods yet. The
ideal is an idea because it exists only in the mind. But as Cicero says,
the ideal, illud quo nihil esse possit praestantius, ‘does shine forth
(eluceat) at some times and in some places, more frequently in some
speakers, more rarely perhaps in others’. This is perhaps a contradic-
tion. But it is this stance, I want to suggest, that brings Cicero into
contact with Longinus.
The Longinian sublime manifests itself in the same erratic and
fitful way as Cicero’s ideal: it is never seen but only glimpsed. Unlike
‘experience in invention and ability to order and arrange material’,
which ‘cannot be detected in single passages’, sublimity, Longinus
writes, ‘produced at the right moment (καιρίως), tears everything
apart like a thunderbolt, and exhibits the orator’s complete power at
a single blow (εὐθὺς ἀθρόαν δύναμιν) and with a single sublime stroke
(ἑνὶ ὕψει)’ (Subl. 1.4; 36.2).6 Invention and arrangement are imper-
ceptible at a glance because they require a long view of the whole.
Sublimity is a quality of style that appears (ἐκφαινομένην) in moments
(ἐκ ἑνὸς ἢ ἐκ δυεῖν); it can only ‘be detected in single passages’. But it
appears so suddenly, so overwhelmingly, and so blindingly as to be
felt more than perceived: to use a Longinian metaphor, ὕψος hides,
so to speak, in its own light, in its very own effulgence (cf. 15.11; 17.2).
5
See Long (1995) for Cicero’s philosophical credentials.
6
Translations of Longinus are from Russell, in Russell and Winterbottom (1972),
with occasional adaptations, as here.

If ὕψος cannot be ‘perceived,’ what are its recognizable features?
And through what faculty can it be ‘detected’? Longinus’ treatise On
the Sublime (περὶ ὕψους) is an attempt to answer these questions. And
it exhibits some of the very same paradoxes that Cicero’s treatise on
the sublime orator does. In fact, in his odd conjoining of Platonism
and oratory, Cicero will help us to identify more closely what it is
about the sublime that is so peculiarly constitutive of it.
Both authors are describing, in effect, a kind of beauty that lies
beyond the senses, but one that, they allow, can only be imagined or
felt through the senses. Longinus will emphasize, for instance, that
sublimity, though it is detectable only at moments, nonetheless
endures beyond the moment by virtue of its impact on the mind
and the memory: what ‘endures only for the moment of hearing
(ἀκοῆς) is not really sublime . . . Real sublimity contains much food
for reflection (ἀναθεώρησις), is difficult or rather impossible to resist,
and makes a strong and ineffaceable impression on the memory’
(7.3). It is a φαντασία (7.1). Sublimity has the quality of an experience
that is anchored in a literary moment (this text here), yet its true
legitimation comes in the repeatability of that experience in a reader’s
mind. The sublime is, at least in its pretensions, textually idiosyncratic
and canonically universal.7 Both aspects are essential to the ever
elusive sublime, which reflects in its core the instabilities of its
momentary anchorage (cf. 2.2).
Sublimity is not something you can hear, nor is it simply a matter
of meaning, because it exists beyond the dimensions of both. ‘Sub-
limity’, Longinus writes, ‘is an echo (ἀπήχημα) of a noble mind.’ And
echoes are both fleeting and uncertain, and an attenuation of sound
(καὶ φωνῆς δίχα). For Longinus, sublimity resides in Ajax’s silence in
the Underworld, in the words he fails to speak (9.1), or in Sappho’s
‘broken tongue’ (10.2), or in the interrupted syntax of Demosthenes’
periods (ὑπερβατά; 18; 22), or else in the larger periodicities of the
physical world that gape terrifyingly and sublimely in the imagination
(9.5–8, 35.3–4). It is its character as ἀπήχημα, as ‘irresistible’ emotion,
shock, and grandeur (πάθος, ἔκπληξις, and μέγεθος), that renders the
Longinian literary sublime, in its irrationality, irreducible to sound or
to meaning. ὕψος has a greatness that is greater than either; and it is
disclosed in the gap, which Longinus never wearies of pointing to,
7
See Porter (2001).
208 James I. Porter

that yawns fitfully between the two orders of sound and meaning. The
idea (ἔννοια) evoked by Ajax is expressed nowhere: it merely illumin-
ates the Homeric passage in which it occurs with a diffuse light. That
light, as Longinus elsewhere more readily concedes, is the light of the
voice (φράσις) that ‘illuminates thought’ (30.2). And well-turned
language, on Longinus’ axiology, is ‘man’s natural instrument’: it
‘penetrates not only the ears but the very soul’ (39.3). With this last
comment, Longinus is getting at the realm of what he calls the
‘conceptions (ἰδέας) of words, thoughts, objects, beauties, and har-
monies’ (39.3), by which he is designating not these things, but their
sublime echo (their ἰδέαι). The terminology is faintly Platonic, but the
application is un-Platonic. And this wavering evocation of Plato is
exactly paralleled in Cicero, as we just saw. Let us look a little closer at
Cicero’s position.
Confronted with the task of drawing up an account of perfect
eloquence, Cicero’s problem is not just the question of how anyone
can have ‘a mental picture (species) of eloquence’ (19), but the
conundrum of how anyone can imagine sound. In setting up a
Platonic ideal of the orator Cicero is establishing a paradox, the
paradox of sound you cannot hear but can only imagine, for the
ideal, he writes, ‘cannot be perceived by the eye or ear, nor by any of
the senses, but we can nevertheless grasp it by the mind and the
imagination’ (quod neque oculis neque auribus neque ullo sensu
percipi potest, cogitatione tamen et mente complectimur, 8). What is
the Platonic idea of sound or of the voice? Perhaps there is such a
thing, although I have to confess I do not quite know what it could
signify except in a pejorative and spectral sense, as in the case of the
‘Protean’ voice that is exemplified by Ion or Gorgias (Ion 541e;
Gorgias, Helen). Of course, a Platonic ideal of rhetoric is by itself
something of a contradiction in terms. One need only think of Plato’s
strictures on the correct speaker (ὁ ὀρθῶς λέγων) in Book 3 of the
Republic: ‘variations’ in the voice of the ‘correct’ speaker ‘are not to be
great’: harmonic and rhythmic modulations, accomplices of mimetic
effects, are to be reduced to a bare minimum, and so on (397b).
Similar restrictions are laid on the musical voice in the same work,
e.g. at 399d, where polyharmonic modalities and modulations are
criticized. The target here, as Mladen Dolar says, is not the voice as
heard but ‘the voice beyond words’: ‘the wind instruments [discussed
by Plato] have the vicious property that they emancipate themselves
from the text, they are substitutes for the voice as the voice beyond

words’.8 It is precisely this separation that is spectrally represented by
the euphonist theories to be considered below. The ‘Protean’ voice of
rhetoric must be rejected. Plato requires a ‘rift (discidium) between
the tongue and the intellect’, as Crassus puts it in De oratore. Cicero
will have none of this Platonic rift, which he calls ‘absurd’ in that
same work: absurdum sane et inutile et reprehendum (3.61).
In contrast to Plato, Cicero’s ideal of oratory is emphatically and
richly vocal. Indeed, only the orator’s utterances are speech in the
proper and full sense of the word (Or. 61–4). We might say that
Cicero gets around Plato’s strictures by turning them inside out.
Instead of letting the Protean voice go, he sets out to capture and
identify it completely (and all the vocis mutationes [55]), in part by
describing it and in part by idealizing it. He makes of the voice a
Platonic Form, the forma eloquentiae (19; cf. 100: ‘we have him now,
Brutus, the man whom we are seeking, but in imagination, not in
actual possession [sed animo non manu]’). And that is a paradox,
leading to a crucial tension in his programme.9 Instead of minimizing
the phenomenal voice and reducing it to its perfect yet abstract form
as Plato does, Cicero goes the other way: his goal is to translate the
phenomenal voice into its perfected ideal without sacrificing any of its
palpable qualities. For Cicero, the logic of perfect oratory is one of
absolute excess, but it is an excess of vocality—its extrapolated ideal.
Faced with Phidias or with verses of Homer, ‘we can, in spite of their
beauty, imagine (cogitare) something more beautiful’. And further:
‘With our ears we catch only the copy’, but ‘with our minds we
conceive the ideal of perfect eloquence’ (10). These copies do not
shed their phenomenal richness. Rather, they point to an even greater
richness than they can ever contain. Cicero is in effect establishing a
Platonic idea (or ideal) of inaudible music, of music you cannot hear,
of sound that exists beyond sound, of what might be called the
sublime form of the voice in all of its phenomenal richness. I want to
dwell a little on this idea and ideal of the sublime voice, not because of
its intrinsic interest alone, but because of the curious fact about
Cicero’s treatise, namely that in spite of its idealizing tendency,
it goes on to devote a good part of itself to the phenomenon of the
voice and its sound, to the radical meaning of eloquentia and of
ῥητορική (and Orator is in this respect unique amongst his rhetorical
8 9
Dolar (2006) 45–6. Cf. Panofsky (1924) 6 and Long (1995) 50–2.
210 James I. Porter

writings). The orator’s mouth is always open, even if it only utters
ideal words.
Cicero is in search of the ideal of the empirical voice. As he himself
acknowledges, we cannot take his search in a literal way: ‘for it is not
an eloquent person (eloquentem) about whom I speak, nor anything
subject to death and decay (neque quicquam mortale et caducum), but
that absolute quality (illud ipsum), the possession of which makes a
man eloquent. And this is nothing but eloquence itself (eloquentia
ipsa), which we can behold only with the mind’s eye’ (101; tr.
adapted).10 The very purity of the object searched for marks it as
Platonic, though the meaning of ‘Platonic’ here has changed consid-
erably. Cicero is in search of the immortal (and sublime) quality of
the voice beyond the voice, that which was the common lure of poets,
rhetors, and literary critics in antiquity, and not simply the musicality
of the voice that would cause Strabo to say that pedestrian prose
writing, denuded of poetic adornments and metre, is like speech that
has stepped down from a certain height (ἀπὸ ὕψους τινός) onto the
ground (1.2.6), but that je ne sais quoi which renders the voice seem
somehow more than itself, what Cicero calls ‘that breath of life (spirit
[us] anim[i]) which usually makes . . . passages seem more impressive
when spoken than when read’ (130).
In pursuing this elusive object, Cicero is inserting himself into a
long-standing tradition that is old as Greek literature. The gap
between the voice and its fullest realization is named as early as
Homer and Hesiod, as Derek Collins has shown. Hesiod distinguishes
between ὄσσα, which is the divine voice, from κλέος, ἀοιδή, or αὐδή,
which are its translations into humanly intelligible language.11 The
familiar Homeric contrast between divine and human names suggest
this kind of distinction already, as does the example of the Sirens who
were mentioned above. Collins makes the further intriguing claim,
which is speculative but just possibly true, that divine ὄσσα has no
sound for humans at all: it is a sound we cannot hear, and a sublime
one at that. A secular, philosophical version of this paradox is found
in Gorgias’ conception (or conceit) of λόγος as a purely theoretical
entity that is without echo or resonance.12 ‘If ossa is the dynamic and
10
Kroll ad loc. notes the Platonic allusion of ‘mind’s eye’.
11
Hes. Th. 10, 22, and 31; cf. Hom.h.Herm 4.443–4; Collins (1999).
12
Porter (1993) 287 with nn. 53 and 289. Cf. Pucci (1977) 28, cit. Collins (1999)
250 n. 24.

powerful sound of voice, a quality that only divine singing can control
as beautiful song, . . . opa, names the pleasing or affecting qualities of
voice, . . . especially the musical sound or tone of [a singer’s] voice.’13
In a word, the poet’s voice, fashioned after divine voice through
afflatus, always promises something more than what you can hear or
comprehend: this ‘more’ is the voice’s sublime quality. The critics
analysed in the present essay study the residual echoes of this quality
in a post-Homeric world. It is this search that brings Cicero into
contact with the κριτικοί before him and with Longinus after him,
that is, with the tradition of sublime criticism. These critics, I am
contending, were among the very few in antiquity who showed
themselves willing to confront this elusive quality of the voice in a
direct, theoretical way. When Cicero says that Demosthenes ‘does not
always satisfy my ears, so greedy and insatiate are they and so often
do they yearn for something vast and boundless (aliquid immensum
inifinitumque)’, he is expressing a desire that is common to all these
theorists of the sublime voice (104; 130).
It is quite likely that Cicero has derived his theory of the voice from
the so-called κριτικοί, or Hellenistic euphonist literary critics known
to us only from Philodemus in large part by way of Crates of Mallos.
Like Crates, they posited the faculty of hearing—‘the ear’—as the
supreme arbiter of literary excellence (hence their name, ‘κριτικοί’).14
At the very least, both Cicero and the euphonists, Crates and his
κριτικοί, are dipping into the same well, as Wilhelm Kroll was the first
to observe.15 Let us rehearse briefly some of the hard evidence for this
connection before considering some of its implications.
CICERO AND THE ΚΡΙΤΙΚΟΙ
Cicero refers to literary critics once, approvingly, as reductive eupho-

nists, according to whom poets vocibus magis quam rebus inserviunt
(68).16 These are in all likelihood the κριτικοί named by Philodemus.
Cicero’s own theory, insofar as it treats of sound, is similarly (that is,
13
Ford (1992) 176. Cf. ibid., 184–95, detailing further sublime aspects of divine
speech evoked through mortal poetry.
14 15
Porter (1995a). Kroll (1907); Pohl (1968).
16
See Maslowski (1978).
212 James I. Porter

reductively) euphonistic. He appears to be translating the Hellenistic
theory into Roman rhetoric, though probably he is not the first.17 Like
the κριτικοί, Cicero values aural pleasure (voluptas aurium) as the
goal of oratory in its character as sound that is heard. And this is
linked for both Cicero and the κριτικοί to the psychagogic effects of
oratory (Or. 203; 168).18 Like the κριτικοί, he credits the ear, not
the mind, with the capacity to judge stylistic qualities (aures sunt
iudices; iudicat enim sensus) (162; 183),19 and these include not only
euphonious qualities but compositional ones at the level of the period
and even at the level of meaning, for these are linked: ‘the ear expects
the words to bind the sentence and its meaning (sententia) together’
(168).20 This is a powerful statement and not at all at odds with
the claim made by some of the κριτικοί that it is the ear and not the
mind that judges meaning. Cicero’s statement merely subordinates
the effects of meaning to the instrumentality of sound, for as Crates
would say, meaning reaches the mind as sound.21 Is meaning, for
Cicero, apprehensible through ‘the mind’s eye’? Perhaps in some
contexts, but not those in which style (φράσις) and the voice are
preeminent, ‘for to present ideas without order and rhythm in the
language is to be speechless (infantia [est]) (236).
Again, as with the κριτικοί, the sense of hearing is in charge of
judging (literally, of ‘knowing’) rhythm, the criterion of which is aural
pleasure (voluptas) and not abstract reason (ratio) (162; 183). Fur-
ther, euphony is one of the functions of the art of joining words
together, σύνθεσις (149), and Cicero knows that μετάθεσις is the
experiment by which the euphonic qualities of σύνθεσις can be
appraised on a sound-based aesthetics (81; 84).22 On the other
hand, euphonic qualities are various, and the κριτικοί appear to
have been inconsistent in their location of these: are aural pleasure
and poetic quality to be found in the μετάθεσις, in the supervenient
euphony, or in the supervenient rhythm?23 The problem was hotly
17
Rightly, Kroll, ad 149: ‘während Cic. hier nur an diesen [sc., an den Klang] denkt’,
in contrast to Theophrastus’ doctrine of ‘aesthetically pleasing words’ (καλὰ ὀνόματα).
18
See e.g. Phld. Poem. 1 Janko cols 109, 133, 159, 161–2, 166, 209. All future
references to On Poems 1 are to Janko’s edition = Janko (2000).
19
Cf. Or. 68, 149, 153, 162, 168, 199, 202–3.
20
Cf. Cic. Nat. D. 2.146 on the role of the ear; and cf. Aristid. 2.248.21 (ap. Kroll ad loc.).
21
Porter (1995a) 93–9. Cf. Porter (1989) 156 nn. 43 and 46.
22
Porter (1995b) 143 n. 138.
23
Porter (1989) 174–5 (‘Appendix A’); Porter (1995a) esp. 100–4.

debated, as Philodemus’ writings indicate (see below). Cicero knows
of this problem, too (Or. 181–2):
Is this rhythmical effect in prose produced solely by rhythm or also by a
certain harmonious arrangement and by the character of the words (vel
compositione quadam vel genere verborum [translating, in effect, ποίᾳ
τινι συνθέσει ἢ ποίᾳ τινι λέξει])? Or does each have its own particular
field, so that rhythm appears in time intervals, harmonious arrange-
ment in sounds, and the character of the words appear as a certain
beauty and embellishment of style? Is harmonious arrangement the
source of all and does this produce rhythm and the so-called figures
and embellishments of style, which, as I have said, the Greeks call
σχήματα? But as a matter of fact these three are not identical [ . . . ].
Harmonious arrangement differs from both because it is wholly con-
cerned with the dignity and charm of words (vocum) [viz., words qua
pleasurable sound; see voce iucundum in the ellipsis above].
Finally, in taking a stand on the apparent debate over analogy and
anomaly, Cicero takes up the cause of anomaly, as the debate has been
misdescribed, or rather (in his own terms), the cause of convention-
ality.24 And like Crates before him, the empirical sense of hearing,
refined through practice (τριβή), is the arbiter of such questions too.25
Analogously, euphonism for both Crates and Cicero is not grounded in
nature, as one might expect (and as the Stoics might be imagined to
argue), but in custom, convention, habit, and art. And so when Cicero
says, ‘I am glad to follow custom which favours the ear’, or that ‘custom,
untaught, is . . . an artificer of sweet sounds’, he is restating this radical
plank of the euphonist program (157; 161; cf. 172). In a nutshell, the core
of Cicero’s theory of sound gives us the core of the doctrines dear to the
κριτικοί and known to us only through Philodemus.26
There are, to be sure, differences. Cicero’s ideal orator is not a
mouthpiece of euphonism, and his treatise is intended as a complete
account of the art of rhetoric: ‘no essential topic will be omitted’ (54).
And so, in Cicero we find inclusions and concessions (say, to mean-
ing) that at first sight are incompatible with the radical euphonism of
24
On the alleged analogy/anomaly controversy, see Blank (1982).
25
Porter (1995a) 93.
26
Being ‘untaught’ means being a ‘natural’ resource, and does not preclude the
conventions of art, instruction habituation, and practice (in the language of the
euphonists: τριβή, τέχνη, and the like). See Or. 51 (p. 215 at n. 32 below); Or. 58
(p. 216 below).
214 James I. Porter

the κριτικοί, which subordinates sense to sound (e.g. Or. 236). So far
as we know, the euphonist critics who value sound over meaning do
not advocate the study of ‘the older logic of Aristotle, or the newer
logic of Chrysippus’ as a foundation to rhetoric or its evaluation, the
way Cicero does—though he is admittedly overstating his case for the
sake of dignifying his ideal: his orator must exhibit ‘greater clarity’
than even a philosopher! (115–17). On the other hand, it isn’t at all
clear that Crates’ definition of the κριτικός eliminates such knowledge
from his own ideal of the ‘perfected art of grammar’ (ἡ ἐντελὴς
γραμματική), which is to say ‘criticism’ (κριτική).27 Cicero’s insistence
on decorum is likewise a convention that radical euphonism dispenses
with more readily.28 But Cicero’s position is no less radical than that of
Dionysius of Halicarnassus (Comp. 11, 38.13–15 or Dem. 47, in both
passages applying musical terminology to the irrational effects of
spoken sound), or than Longinus’, who holds that sublime effects
‘depend as much on the harmony [viz., σύνθεσις] as on the thought
[i.e., “meaning”]’ (39.3). On the other hand, when it comes to effects of
sound, decorum simply means whatever is appropriate to and condu-
cive to aural pleasure or other aesthetic effects, and on this score we
find complete agreement among the κριτικοί too.
Even here, some of the apparent differences evaporate on closer
inspection. ‘No word has force apart from a thing’ Cicero says at one
point, but then he quickly adds that the rejection or approval of a
word crucially depends on the way it is expressed (72). It isn’t clear
that a κριτικός, apart from Heracleodorus, wouldn’t agree with either
half of this statement. Certainly Crates would: it is ‘not the sound
itself without the sense’ but the two ‘combined’ that enchant a reader,
even if sound is aesthetically preeminent. Meanwhile, allegorical
meanings have a force that is all their own; they simply coexist with
sounds on a distinct level of appreciation that may nonetheless be
enhanced by euphonic considerations.29 Still, Cicero typically shows a
casualness towards meaning, whose criterial value he readily acknow-
ledges, unlike the κριτικοί, who do not, at least in our experience of
them. Thus, ‘the individual word wins approval which has the best
sound (optime sonat) or best expresses the idea’ (80). Similarly, ‘the
27
Sext. M 1.44, cf. 79: ‘the whole range of linguistic science (λογικὴ ἐπιστήμη)’;
248: κριτική.
28
Cf. Phld. Poem. 1 col. 46.
29
Porter (1995a) 96–8; Porter (1992) 111–13.

audience notice these two things and find them pleasurable—the
words and ideas’ (197). But even here, he follows up this last thought
with another worthy of Crates (but also paralleled in Longinus):
attention to ‘words and ideas’ causes inattention to rhythm; and
rhythm intensifies the pleasure taken in words and ideas (197–8).
This is a corollary of the theme, familiar from Crates, of ‘distraction’
or τὸ ἀποσπᾶν: sound distracts from sense, typically by isolating the
sound and focusing all of one’s aesthetic attention, though in some
cases distraction can enhance the apprehension of meaning in such a
way that permits us to say, not that we understand meaning, but that
we ‘hear’ how meaning sounds.30
Cicero can make such claims because in the last analysis ‘the whole
essence of oratory is to embellish (illuminare) meaning’ (136; cf. 44
and 236). Embellishment takes precedence: Cicero’s values can be
read off the relative weight that he gives to his topics, which is visible
in the amount of space that he allots to each. His discussion of
sonority (elocutio) takes up the greater bulk, a good three-quarters,
of Orator. The first two parts of oratory (invention and arrangement
of subject matter) he dispatches in a few sections, eagerly turning to
questions of style and expression understood in their relation to sound31
(the former ‘require less art and labour,’ 51),32 thus replicating the biases
of the κριτικοί and, as we saw earlier, of Longinus too. It is only natural
that whenever due attention is paid to musical embellishments and
for as long as it is, the aural dimension tends to subordinate meaning,
morals, and other considerations: it points to a purely aesthetic
dimension. For as Dionysius says, in defiance of Plato (Rep. 3.398d9;
400d11–e1), ‘music requires that the words should be subordinate to
the melody and not the melody to the words’ (Comp. 11, 41.18–19).
This claim is not as radical as the thesis that sound powerfully
distracts from the sense, but neither is it incompatible with it.
Some of the differences between Cicero and the κριτικοί are to be
explained by differences in object: the κριτικοί are mainly literary
critics (it is in this context that he names them, if that is what he is
doing), though they frequently extend their principles to oratory,
which they then tend to treat as a kind of poetry, as will Cicero and
30
Porter (1995a) 94–5. Cf. Subl. 41.2: ‘ . . . just as songs distract an audience from
the action and compel attention for themselves (ἀπὸ τοῦ πράγματος ἀφέλκει καὶ ἐφ’
αὑτὰ βιάζεται)’.
31 32
Or. 162; cf. 149, and passim. Cf. Dion. Hal. Thuc. 34.
216 James I. Porter

Dionysius of Halicarnassus later.33 What is remarkable, on the other
hand, is the extent to which Cicero integrates the lessons of literary
euphonist criticism into his views on rhetorical theory, effectively
bracketing and suspending the issue of subject matter. The reason
is to be found in the common starting-point of their inquiries: ‘the
voice possesses a marvellous quality (mira natura)’ (57); it has a
musicality to it (‘even in speech, there is a sort of singing [cantus]’):
‘therefore let art follow the leadership (ducem) of nature in pleasing
the ear’ (58).
We can overhear in this last claim Crates’ similar claim about the
ἡγεμονία of sounds in critical judgement.34 ‘Certainly natural excel-
lence of voice is to be desired’, Cicero holds in protest against Plato,
and ‘the superior orator will therefore vary and modulate his voice;
now raising and now lowering it, he will run through the whole scale
of tones’ (59). Some of this is a commonplace in rhetorical discourse:
even Aeschines discusses ὁ τόνος τῆς φωνῆς in Demosthenes, meaning
Demosthenes’ use of pitches.35 What the antiquity of the topos points
to are the common sources of literary and rhetorical criticism in the
ancient theory of music and of poetry as music, which is to say the
reciprocal influences of musical theory on poetic theory and vice
versa, a point that has been made in the past but which is too easily
neglected, though the tide is slowly turning.36 We would do well to
recall Westphal’s prophetic reminder from 1861: ‘A further reason for
the neglect of rhythm [in modern scholarship] lies in the erroneous
notion that the writings of the ancient rhythmicists are primarily
33
Phld. Poem. 1 col. 199, where Heracleodorus treats Demosthenes, Xenophon,
and Herodotus as poets, not as prose writers. Cf. Nardelli (1983) 108. Cic. Or. 67 (on
Plato and Democritus) and 37–8 and 183 (for the general principle); Dion. Hal. Comp.
24, 121.10–21; Comp. 26, 135.20–136.4. (All references to Dionysius are to Usener–
Radermacher’s edition.)
34
Phld. Poem. 5 col. 16.5 Mangoni; Poem. 1 cols 114.14 and 132.19; Poem. 2, P.
Herc. 994 col. 21.23 N = Tr. A col. 21 Sbordone; cf. 994 col. 25.4 N = Tr. A col. 25
Sbordone. References to treatises (‘Tr.’) A, B, and C are to Sbordone (1976), with
Arabic numerals replacing Sbordone’s Roman numerals. See Westphal (1861) 21 on
the figure of the ἡγεμών, or the ‘tactangebender Musikdirector’.
35
In Ctesiph. 209–10. A slow evolution in the poetical and musical tradition is
suggested by later allusions to, e.g. Pratinas, who speaks of the ‘tense’ and the ‘relaxed’
ἁρμονία (Ath. 624f).
36
Kroll (1907); Costil (1949); Pohl (1968); Porter (1989) 174 n. 139; Janko (2000)
173. Interest in aural aesthetics in classics and elsewhere is resurgent. See the essays by
Page and Dillon in Jaeger, ed. (2010); Dillon (2012); Butler (2015); Steiner (2015);
Gurd (2016); Butler and Nooter, eds (2017).

directed to music, and not to the rhythm of poetry as well.’37 To
underscore the point once again, it is the insistence on the musicality
of language that is the driving inspiration of the euphonist analysis
of poetry and of rhetoric, and which brings both to the brink of
the sublime. The sublime, I want to suggest, names what is so
problematic—but also completely defensible—about this insistence
on the irrational aspects and effects of language.
Were there space, we would want to discuss the various sources
shared by Cicero and the κριτικοί, starting with Hieronymus of
Rhodes (fl. mid-third century CE), whom Cicero credits with the
view that ‘there is rhythm in prose’, the same as ‘those of poetry’
(Or. 180; 166–7; 185),38 and leading back to the earliest theories of the
literary voice in Aristoxenus (whose rhythmic theory Cicero follows),
and even prior to Aristotle (for example, Gorgias’ pupil Licymnius),
for instance the ἁρμονικοί mentioned by Plato, Aristotle, and Aris-
toxenus.39 And we would want to consider what the κριτικοί made of
rhythm (a subject about which we know too little).40 We know that
the κριτικοί made claims about εὐρυθμία parallel to their claims
about euphony; they operated with a notion of the σύνθεσις of
rhythms (σύνθεσις τῶν ῥυθμῶν); and they credited the ear with the
capacity to judge the particulars (κατὰ μέρος) of rhythmic shapes.41
Perhaps rhythm was treated in a missing section of On Poems.42 This
loss, if it is one, is greatly to be rued, because here we would undoubt-
edly have had a direct link to the work of the musicologists of the fifth
and fourth centuries.43 The connection to rhythm is a natural one; the
later critics are after all trying to capture the ‘music’ of language, what
37
Westphal (1861), 3; emphasis in original.
38
On Hieronymus, see Porter (2010) 339–41.
39
Arist. Rh. 3.2, 1405b6–8 with Porter (2010) 314–15. Arist. Pol. 8.5, 1339b18
shows that music theory had already been applied to the analysis of literary style
(Kroll (1907) 93). On the fifth-century ἁρμονικοί, see Plat. Rep. 531a–b; Arist. Po.
An. 78b–79a.
40
See Phld. Poem. 1 col. 101 (a glimpse only).
41
Phld. Poem. 2, P. Herc. 1676 col. 3.2 = Tr. C col. 14 Sbordone; P.Herc. 994 fr.
19.7–13 N = Tr. A col. b Sbordone; Phld. Poem. 1 col. 101.2–8 Janko.
42
Cf. Phld. Poem. 1 cols 101 and 151 Janko; Poem. 5 col. 27.2 Mangoni.
43
Heraclides of Pontus, treated by Philodemus, seems to be a mediator of fifth-
century views of musicality in poetry (see n. 46 on Hippias) collected by Glaucus of
Rhegium ([Plut.] De mus. 1131F and 1132E). Heraclides correlated musical and
literary styles in his lost treatise, On Music (fr. 163 Wehrli). See Porter (1989) 166
at n. 94 and Janko (2000) 134–8 for discussion.
218 James I. Porter

Dionysius calls, in a nice συμπλοκή, τὸ τῆς μέλος φωνῆς.44 Still, given
that the music of language is not exactly a property of language,
because it is a quality that lies beyond language, we can never expect
anything more than analogies, which is all that we in fact get.45
Aristoxenus gives us the first clear attestation of such an analogy
between the two kinds of σύνθεσις, musical and linguistic, although he
was probably preceded by Hippias and others.46 I would like to close
by looking at one such analogy in Cicero, which will bring us back to
Longinus and help to clarify the notion, which no doubt is still
mysterious, of sound you cannot hear.
In section 187, Cicero writes the following: ‘If one passage is
constrained and choppy and another is diffuse and flowing, this
cannot proceed from the nature of the letters [i.e., from their sound],
but from the varied arrangement of long and short intervals (inter-
vallorum); and since prose, which is an intertwining and blending of
these intervals, is at times sedate, and at times rapid, such a phenom-
enon as this must depend on rhythm.’ How does one hear rhythm? In
rhetorical theory, intervals can variously stand for differences in pitch
and tone, for differences in time, or for silent pauses.47 Intervalla in
44
Dion. Hal. Comp. 11, 43.5–6. Cf. Dion. Hal. Dem. 48, 233.1–2 on tone and time
(μέλος καὶ χρόνος) as generally characterizing whatever lies musically beyond the parts
of speech.
45
See e.g. Dem. 2: ‘The style of Lysias . . . bears the same relation to that of
Thucydides as the lowest to the highest note on the musical scale.’ Cf. Comp. 11,
40.8–16.
46
Aristox. Harm. 27.18–20 compares the construction of melody with the synthe-
sis of letter-sounds in language, a striking predecessor to the κριτικοί. See Dion. Hal.
Comp. 11, passim, and esp. pp. 40.17–41.1 On Hippias, see DK 86A11 and 12 (Hippias
was the first to inquire περί τε γραμμάτων δυνάμεως καὶ συλλαβῶν καὶ ῥυθμῶν καὶ
ἁρμονιῶν). Isoc. Antid. 46–7 is explicit about the musicality of speech-making.
47
intervalla, distinctio, et vocis genera permulta (Cic. Nat. D. 2.146), rendered by
Pease (ad loc.) as ‘differences of pitch . . . ; representing Greek διαστήματα’. Hubbell
oscillates between ‘differences of tone, of pitch, and of key’; and Pease adduces De or.
3.186 (quod intervallis distinguuntur), where ‘distinctio is apparently used of rhythm’,
but there is disagreement as to the meaning of distinctio and intervallum there too.
Strangely neglected is a further meaning of ‘pause’. Cf. Dion. Hal. Dem. 38,
210.23–211.4, where temporal duration (χρόνος) is in the first instance not rhythmic
but phonic (a gap of silence, a pause) that is filled in ‘by musical and metrical writers’
with semivowels so as to avoid harsh effects of hiatus and by rhetors with a pause
(cf. interpuncta intervalla, morae respirationesque delectant at Or. 53). Of course,
‘pause’ (ἀνάπαυσις) could signify both silence and rhythmical gap-fillers in rhythmical
theory (see e.g. Hermog. Id. 2, 259.25–260.3 Rabe, cit. by Kroll (1913) ad Or. 181).
Finally, ‘silences’ occur between any two sounds that cannot be phonically blended,
such as ν and χ in ἐν χορόν (Comp. 22, 101.12–21). Typically, the focus is not on letters

the present passage refers, however, to the χρόνοι, or quantities of
time, of Aristoxenian rhythmic theory, whose proportional relations
constitute rhythmical relations, and which (as Cicero says) have
nothing to do with metrical feet (193–4).48 The point here is that
the ear, when it attends to rhythm, listens not to the absolute sound of
letters or even of syllables but to proportions of time designated by
the rhythmic pulses of arsis and thesis, that is, to relations.49 Two
features on this theory stand out: on the one hand, the minimal,
indivisible character of the intervals (detectable and measurable only
by the ear), a model with clear parallels in harmonic theory (τὸ
ἐλάχιστον τῶν συμφώνων διαστημάτων) and in its close relative, the
phonic differences of euphonic literary theory (φυσικαὶ διαφοραί τῶν
φώνων),50 and on the other what we may call the epiphenomenal or
supervenient character of rhythm: like euphony, which supervenes on
synthesis, rhythm emerges from the interplay of these quasi-abstract
relational elements (durations of time) in varied proportions and
tempos (viz., intervals), as an appearance (φαντασία) in the mind of
the auditor.51 And yet, what intervals signify, in rhetorical theory at
least, are a threat of rupture.52 And it is this play with gaps—first
(a poor rendering of στοιχεῖα) so much as it is on sounds that are characterized by

δυνάμεις, viz. the chromatic elements of the voice; periods, in turn, are merely
punctuations of the breath (πνεῦμα) (Comp. 22, 110.21–111.1).
48
The distinction (mentioned at Comp. 17, 68.7–12) is pre-Aristoxenian. See
Norden (1971 [1909–18]), 1:53 on this ‘law’ of rhythmic prose: ‘die Rede darf nie
metrisch, muß aber immer rhythmisch sein’.
49
A point that Theophrastus would try to palliate. See Barker (1989) 111; and
Theophrastus ap. Porph. In Harm. (Barker, (1989), 113 and 117). Cf. Or. 67 on
measuring rhythmical qualities by the ear, in the way that Aristoxenus prescribed.
50
Aristox. Harm. 20.7–8; Phld. Poem. 5, col. 27.20–9 Mangoni (Crates); Dion. Hal.
Comp. 14, 49.1. The two terms διάστημα and διαφορά are conjoined in Polyb. 38.5.
51
The language is remarkably stable over the centuries. See Aristox. Harm. 8.23
(κατά γε τὴν τῆς αἰσθήσεως φαντασίαν), 9.2–3; 10.11–21; 48.22, with Barker (2007),
143–4; Dion. Hal. Comp. 22, 110.8–9 ( . . . οὐκ ἐῶντα τὴν ἀκρόασιν ἑνὸς κώλου
συνεχοῦς λαβεῖν φαντασίαν); Porph. In Harm. 31.18–21 Düring: ‘The notes, because
they are closely successive [though bounded by ‘gaps’ (διαστήματα) that are “small
and cannot be apprehended (ἀκατάληπτα)”] create the appearance of a single sound
(ἑνὸς ἤχου ποιοῦνται φαντασίαν) stretched out over some amount of time’ (tr. Barker
(1989), 236). The language also appears in atomistic accounts of perception, e.g.
Epicurus, Ep. Hdt. 49; Lucr. DRN 4.794–8. The underlying idea goes back to the
Presocratics’ theories of perception. Cf. Theophr. Sens. 63 (Democritus); on Empedo-
cles, see Porter (2010) 152–4. Further, Richter (1974) 279 with n. 58.
52
Comp. 25, 109.14–20: juxtaposed vowels ‘interrupt’ the flow of sound, ‘whereas
it is continuous and smoothly blended sounds that produce an euphonious effect’.
220 James I. Porter

among the elements of time or sound, for these are distinguished
relatively and not absolutely, and then between these and their
synthetic perception—that opens up the prospect of larger aesthetic
effects. The point is that when you look for the sources of sound, what
you find is a series of gaps. To put this in its most provocative form,
we might say that sound is an arrangement of silent gaps.
GAPS
Longinus will make much of such gaps. As he writes about three

verses from Euripides’ lost Antiope, while establishing the critical
remark that ‘it is in the σύνθεσις rather than in the sense that
Euripides’ greatness appears,’ ‘the words are propped up by one
another (they are ἐξερείσματα) and rest on the intervals (τῶν χρόνων)
between them; set wide apart like that, they give the impression of
solid strength’ (40.3–4; tr. Russell).53 As Russell glosses the meaning
of σύνθεσις here, what is meant is ‘putting the words together in an
effective and euphonious order’, that is, their ἁρμονία. The synaes-
thesia of the visual metaphor, which recalls Dionysius of Halicarnas-
sus,54 not only reflects the passage beyond linguistic perception of the
sort we have been discussing throughout. It also rests on the transla-
tion of voice into time (‘producing on the hearer the effect not of
speech but of rhythm’) and then into space (41.2).55 Let us simply
emphasize the concept of ‘impression’ and further ask ourselves how
intervals sound. Such intervals, architectonic spacings ‘that give rise
53
Literally, ‘with a view to,’ hence: ‘for the sake of creating the appearance of solid
strength [or “grandeur”]’ (πρὸς ἑδραῖον . . . μέγεθος).
54
Comp. 22, 96.10–22 and esp. 20, 91.3–92.3 (see Porter (2016) 407). Cf. Dion.
Hal. Dem. 38, 210.9–14. An ancestor for the construction-metaphor: Phld. Poem. 2, P.
Herc. 994 cols 34–6 N = Tr. A cols 34–6 Sbordone (where εὐπαγές again figures). Such
analysis, in addition to playing with (and estranging) the perceptual modes of
aesthetic apprehension, lays emphasis on the constructed quality of the object that
is being simultaneously presented as seamless. The passage from natural sound to art
and technique (see p. 220–4) is already contained in the figure of built dimensions,
indicating that no passage ever really occurs.
55
The spatial and temporal dislocations can take on a further significance, namely
as a reminder of the pastness and of the (broken) monumentality of the poems under
consideration. This is the classicistic element of the euphonist agenda, on which see
Porter (2005b).

to a delay (ἀναβολή) and interruption (ἐγκοπή) of the rhythm’, are for
Dionysius of Halicarnassus the hallmark of the austere style, for
instance in Pindar and Aeschylus (Comp. 22). The sublime, by con-
trast, is not reducible to a style, for it is the quintessence of every style.
It is their common impossibility.
Take the example of Dionysius of Halicarnassus. His ideal is a
synthesis of austere and smooth styles, virtually acknowledged to be
impossible.56 But then so are all three kinds of synthesis: these are
ideal, never purely instantiated (Dem. 37)—as is euphony itself. The
point is rarely grasped, but Dionysius is remarkably explicit about this
fact. In chapter 18 of On the Composition of Words, Dionysius speaks
of rhythmic purity in its ideal form (κατ’ εὐχήν): what he means is
that pure and unadulterated euphony is a desideratum that is never
met by language.57 Elsewhere, the emphasis is on the artificial
disappearance (ἀφανίζειν) of natural dysphony and the artificial
appearance (φαίνεσθαι) of euphony (Comp. 18, 73.21; 16.67.3–14;
26.136.21).58 The point is that euphonic σύνθεσις is in actual terms a
pis aller relative to its ideal goal of perfect euphony. As a consequence,
the ear must be deceived and cheated into an illusion or intimation of
(impossible) euphony.59 It is this recognition that gives the concept of
οἰκονομία, the technical management of the appearances and effects
of language, its fullest meaning.60 The sublime is the height of this
deception.
Dionysius is a special case who deserves to be treated separately.
But where he shows himself to be a sublime critic is to be found in his
most characteristic and culminating critical gesture: it is in his hunt
for concealed phonic features, for instance incomplete metrical pat-
terns, missing syllables and tempi, disguised resolutions, and the like,
made so by writers who ‘cause us to forget the metre’ and ‘prevent
56
Cf. Dem. 37, 209.14–210.1; Dem. 50, 236.19–21; and Comp. 21, 95.14–96.1 (etc.)
with Porter (2016) 220–3.
57
Styles have no pure existence; they are always mixed, and hence named after a
‘dominant’. So, e.g. Dion. Hal. Dem. 37, 209.14–210.1. This is true for Demetrius as
well. See Porter (2016) 221–4; 259.
58
Cf. Comp. 16, 67.3–4: ‘we must try to cover up (ἀφανίζειν) the natural defects of
the inferior letters by interweaving, mixing and juxtaposing’, etc.
59
Cf. on the smooth style, see Comp. 23, 111.18–112.17 and 113.16 on its quality of
being ‘deceptive’ (ἀπατηλός), for the same reasons as outlined here.
60
Comp. 18, 74.3, where the functions of οἰκονομεῖν (‘arranging,’ ‘managing’
compositional qualities) and διακλέπτειν (‘cheating’ or ‘deceiving’ the ear) are linked.
222 James I. Porter

us from recognizing’ the rhythm (Comp. 26, 136–7),61 and which
Dionysius’ sharp ear can detect like few others and his pen can supply
again, in a resounding demonstration of his point—that the music of
language lies just below the threshold of what you can hear (Comp.
25). Cicero attests to the precedence of this critical manoeuvre.62 But
more broadly we might say that the mind’s—or is it the ear’s?—
‘completion’ of missing sound is emblematic of the euphonistic
phenomenon in general. At least in speech, which ‘has no measures
of rhythmical intervals [viz., beats] like those given by the piper [in
poetry], the whole periodic form (species) of the sentence is rounded
out and brought to a finish in a way which can be judged only by the
pleasure of the ear’ (Or. 198). The sound is never in itself complete; it
is ‘completed’ in a species—an image—that exists only for a listening
subject, one who strains after inaudible harmonies. It is in this
potential for a missing element that listening becomes sublime and
that rhetoric becomes (in Cicero’s words) ideal.
The essence of euphony for Crates and the κριτικοί is likewise
sublime, but with an interesting variation. For them, the pursuit of
the unrivalled excellence of sound moves in two directions at once.
The one is the search for pure sound, ‘the sounds themselves’ (περὶ
τῶν ἤχων αὐτῶν), which is to say, the purity of sound itself in its
material and phenomenal character (Poem. 1, col. 83). Literally taken,
this route leads to a focus on φωνήεντα, that is, on self-sounding
vowels, ‘for nothing is supported upon these (ἀπερεί|[δεται]) but the
sound itself (αὐτὸς ὁ ἦχος)’ (col. 84.4–6). The radical autonomization
of sound, beyond even the notion of ‘simple sound’ (ἁπλῆ φωνή),63
beyond even the separation of sounds from their attached mean-
ings,64 and (optimally) involving the fewest physical strictures of
the mouth (as Dionysius shows),65 produces not only striking
61
Cf. Comp. 25, 127.10–11: ‘ . . . has concealed its identity (πεποίηκεν αὐτὸ
ἄσημον)’. said of a resolved elision that disguises ‘a complete elegiac pentameter’.
62
Or. 67: ‘For everything which can be measured by the ear, even if it does not make a
complete verse—that is certainly a fault in prose—is called rhythm.’ This shows that
Dionysius’ practice is inherited. A collection of trimeters from Isocrates’ speeches
(Spengel (1828) 153) shows that the practice was widespread; see Kroll (1913) ad Or. 190.
63
Phld. Poem. 2, P.Herc. 994 fr. 20.25 N = Tr. A col. c Sbordone. See at n. 67 below.
64
Phld. Poem. 1 col. 114.10–15 and 115.1–3.
65
Comp. 14, 50.12–52.13: vowels when long are the most attractive ‘because they are
sounded for a long time’; short vowels are inferior to long vowels ‘because they lack
volume (ὅτι μικρόφωνά τε ἐστί) and restrict the sound’; alpha is most open and least
supported by the structure of the mouth; upsilon is the second to least euphonic of

architectural imagery, which underscores the limits to which euphonist
description takes itself, but pure pleasure, and even ecstasy: ‘it compels
our nature to be exalted (μεταιω[ρεῖσθαι)’.66 But this is the utopic
domain of sonic fantasy. Pure sound as such, vocalic sound, does not
exist in language. And so, ‘if we add another of the letters’ to the pure
sound of a vowel ‘difference’ (διαφορά) invariably sets in; the mouth has
to move and change its shape; gaps (interruptions) intervene.67 The
analytical pursuit of euphony is in fact a recognition of such differ-
ences in sonority, blent to a sphere of consonance on another level.
The reason behind this insight is the recognition that language is the
contingent effect of letters in synthesis: euphony lies beyond the reach
of reality; it is an effect that paradoxically disguises its own impossi-
bility. This insight, incidentally, is Platonic, but not exclusively so.68
Hence the second direction of the pursuit, which is holistic and
systematic rather than vocalic: this is the study of sounds in combin-
ation, in euphonic clusters. What is recovered on this approach is the
system of sounds in their joint complexity; but what is lost is precisely
the vocalic purity of sound.69 ‘Just as in an instrument, and even more
so in a bow, [so too] there will be no harmony in the whole (τ[οῖς]
ὅλοις)’ unless there is ‘relaxation and tension’ (so Crates, Phld. Poem.
1 cols 93–4). Crates’ theory of στοιχεῖα on this second approach—and
Crates generally seems to have theorized the teachings of his eupho-
nist predecessors in a rigorous and novel way—is a view about the
relations that obtain among the letters and the sound-effects thereby
produced ‘in the whole’. But we can go further, for his theory,
I believe, is about the difference between letters as στοιχεῖα (strictly,
these vowels: the lips contract, the sound is choked and made ‘thin’; iota is last, most
restricting ‘the column of breath’ (τὸν αὐλὸν τοῦ πνεύματος), etc.
66
Phld. Poem. 1 col. 84.17–20 (Pausimachus, a κριτικός).
67
Poem. 1 col. 84.7–9; Dion. Hal. Comp. 22, 104.3–6: ‘The process of the mouth’s
altering from one shape to another that is neither akin to it nor like it entails a lapse of
time, during which [or “by which” (Usener)] the smoothness and euphony of the
arrangement is interrupted (διίσταται).’
68
‘Audible sounds which are smooth and clear, and deliver a single series of pure
notes (τὰς ἕν τι καθαρὸν ἱείσας μέλος), are beautiful not relatively to something else,
but in themselves, and they are attended by pleasures implicit in themselves’ (Phlb.
51d6–9; tr. Hackforth). Plato is doubtless re-characterizing prior euphonist theory,
not innovating. He is also performing a kind of reductio ad absurdum, by reducing the
idea of absolute sound to an inaudible extreme: sound becomes, per impossibile, its
own Form. See Porter (2010), 88–9.
69
Hence the recognition that all three kinds of synthesis are ideal, never purely
instantiated; see at n. 56 and Pohl (1968) 151.
224 James I. Porter

these have no proper sound; rather, they represent phonic possibil-
ities, much like phonemes) and letters as sounded in discourse and
organized in poems and poem-like sentences, whereby phonemes are
‘realized’ (ἀποτελεῖσθαι).70 And yet, this second approach to euphony
is likewise productive not merely of pleasure but of ecstasy, through a
kind of Longinian compulsion (ἐνθουσιῶ; [ἐ]πικρ[α|τ]ῇ ἡμῶν).71
Whence does the pleasure come? That is the mystery of (irrational)
sound beyond the rational system of sound. Or else—to turn this the
other way around—it is the search for the hidden reasons (αἴτιαι), the
art and artistry, within the irrational effects of sound.72 The sublimity
of euphonic criticism lies precisely in this complementarity that
never quite adds up. It is, so to speak, the ‘gap’ at the heart of this
critical system.73
It never adds up because, we might say, sublimely born sound
(εὐγενὴς φωνή)74 borrows from the features of material sound even
as it exceeds those features. For all its phenomenal immediacy,
euphony is harder to trace back to its hidden causes than rhythm,
which is at least patterned and mathematically so. The specificity
of sound is elusively of the moment and ‘punctual’ (this sound here); it
is grounded in its material coordinates but distributed over their σύνθ-
εσις. Euphonic criticism is thus a science of the impossible; the theory
is so highly staked upon the ‘absolutism’ of the poetic particular (the
view that sound is ἴδιον, proper to its location, not κοινόν, or shared)
that it becomes difficult to compare particular sounds with one
another, let alone to discern their exact reach.75 Dionysius’ criticisms
range over whole strophes at times, but he can also treat a ‘combined’
70
This is why there are disputes in antiquity as to just how many letters of the
alphabet (viz., primary sounds) there are. See Comp. 14, 50.1–11; Sext. M 1.11; schol.
Dion. Thrax, xl.6–8; xl.22; 32.25–33.5; 474.20–3; 496.17–18 Hilgard; and Porter
(1989) 177, where the soundlessness of the ‘elements’ of euphonism is discussed—a
forerunner to some of the thesis of the present essay.
71
[ἐ]πικρ[α|τ]ῇ ἡμῶν: p. 225 below; ἐνθουσιῶ: Poem. 2, P.Herc. 994 col. 7.8 N = Tr.
A col. 7 Sbordone; Phld. Poem. 1 cols 158.21 and 162.13 for θεάζειν and ἐνθεάζειν,
verbs which justify Philodemus’ calling his opponents Corybants in the hallowed
tradition of Plato. See Porter (2016) 239–46 for a fuller discussion.
72
See Phld. Poem. 1 col. 123.22–7.
73
See Porter (1995b) 113; cf. Pohl (1968) 137.
74
The phrase makes a phonetic pun on εὐφωνία, as Asmis (1992) 167 rightly
points out.
75
See n. 41.

sound as an ἰδία φωνή.76 This leads to the problem of generalizing
over cases (how can you compare two or more sounds if each lays an
absolute claim on our attention?) and to the bias against rules
(θέματα), which Cicero also resists (Or. 36; 43; 237), as does Longi-
nus. On their view, there can be no science that governs absolute
particulars. But the same insight also leads to a further problem—
namely, that to pursue sound in its purity, as rigorously as the
euphonist critics do, is nearly fatal to the object pursued. This danger
affects all of sublime criticism, and is its ultimate allure.
There are other, more immediate but also more superficial con-
nections to be made between the κριτικοί and the Longinian sublime.
Sounds dominate us by arousing us to exaltation (ἐ]πικρ[α|τ]ῇ ἡμῶν,
εἰς ⌞τ⌟ὴ̣ ν ἐπι⌞π⌟ρ̣έπειαν ἀνακ[ι]νεῖν)—so Crates (Poem. 1, col.
114.16–17).77 Similarly Longinus, when he treats the fifth source of
sublimity, σύνθεσις (‘ἁρμονία’), which encompasses ‘the combination
and variety of sounds’ and which can suggest (κινοῦσαν), if not
produce, various ideas in the mind as well as effects of beauty and
melody, ‘charming’ the hearer and holding complete domination over
her mind (ἡμῶν τῆς διανοίας ἐπικρατοῦσαν, 39.3).78 Crates’ view that
poetry must use the emotions as allies (τοῖς δὲ] π‹ά›θεσιν (Hausrath:
τοῖς δ’] ἤθεσιν Janko) συμμά|[χοις χρ]ῆσθαι τὴν πόη|σιν) (Poem. 1 col.
132.21–3), where ‘emotions’ are elsewhere clearly marked as psycha-
gogic,79 further recalls Longinus’ view that ‘figures are natural allies
(συμμαχεῖ) of sublimity and themselves profit wonderfully from the
alliance (ἀντισυμμαχεῖται),’ and that ‘sublimity and emotion are a
defence and a marvellous aid (ἐπικουρία)’ in rhetorical writing
(17.1–3).80 But ultimately the hallmarks of the sublime are, as
Longinus himself says and shows, elusive and hard to name. It is
76
Comp. 14, 53.2–3: the simple sound of zeta is combined (σύνθετα) from sigma
and delta.
77
See Janko ad loc. for discussion.
78
There may an echo with Heracleodorus here as well, another κριτικός who held
that only language treated with artifice (τὰ πεποημένα) ‘moves’ the hearer, not
language without artifice (τὰ οὐ πόητα) (Phld. Poem. 2, P.Herc. 1081 fr. 23 N = Tr.
C fr. n Sbordone). See Porter (1992) for Crates’ place in the sublime tradition, which
in literary criticism is invested in the sublimity of art, not of randomly occurring
sounds.
79
Poem. 1 col. 136.14–20 (where both character and emotion seem to surface,
possibly in a contrast).
80
Porter (1995b) 147, cf. 110.
226 James I. Porter

above all the confrontation with elusiveness that marks the presence
of the sublime.
Crates and the κριτικοί resemble the curious and eponymous
character from a tale by E. T. A. Hoffmann, Councilor Krespel (‘Rat
Krespel’), who is in search of the perfect sound. He builds violins that
he plays only once and then lays aside (all for the sake of relishing the
pure and original immediacy of their sound, which is by definition
unrepeatable), and he acquires others only to dissect them for what
they conceal within:
I am totally convinced that something special lies within the inner
structure [of the violin], and that if I took it apart into its constituent
pieces it would reveal a secret that I have long been in search of.81
The thing (the instrument, the structure) is ‘dead’ in itself (dies tote
Ding), like the letters of the Greek alphabet described by Antiphanes
in the fourth century as universally aphona, ‘voiceless’.82 Plato, too,
knows the distinction, for instance in the Phaedo:
One might make the same argument about harmony, lyre and strings,
that a harmony is something invisible, without body (ἀσώματον), beau-
tiful (πάγκαλον) and divine in the attuned lyre, whereas the lyre itself
and its strings are physical (σώματα), bodily (σωματοειδῆ), composite
(σύνθετα), earthy and akin to what is mortal. Then if someone breaks
the lyre, cuts or breaks the strings and then insists . . . that the harmony
must still exist and is not destroyed because it would be impossible
for . . . the harmony . . . to be destroyed before that which is mortal, he
would say that the harmony itself still must exist and that the wood and
the strings must rot before the harmony can suffer (τι παθεῖν).
(85e–86a; tr. Grube)
Plato’s predicament is not dissimilar to Krespel’s, or to that of the
euphonists for that matter. He knows that harmony (musical beauty)
is an epiphenomenon of a corporeal event, even if harmony also
transcends the body that gives rise to it initially, and in doing so it
seems to approach the status of the divine. But the nature of aural
decay, so to say, renders Plato suspicious of the analogy from which
this brief glimpse of musical harmony emerges, namely that between
81
‘Ganz überzeugt bin ich, daß in der innern Struktur etwas Besonderes liegt und
daß, wenn ich sie [sc., die Violine] zerlegte, sich mir ein Geheimnis erschließen würde,
dem ich längst nachspürte’ (‘Rat Krespel’, in Hoffmann (2001) 48).
82
Antiphanes, Sappho fr. 196 K-A.

the soul and musical harmony. Socrates and his interlocutors ultim-
ately decide that the soul is not a harmony (95a), owing to the
epiphenomenalism that such an analogy presupposes: musical har-
monies supervene on physical parts and are themselves constituted of
those parts; the soul must be qualitatively and causally distinct from
the mortal parts in which it provisionally takes up residence. ‘Does
not the nature of each harmony depend on the way it has been
harmonized?,’ Socrates asks. And ‘do you think it natural for a
harmony, or any other composite, to be in a different state from that
of the elements of which it is composed?’ (93b; 93a).
In insisting upon the composite (συγκειμένη) nature of musical
harmony (92b1),83 Plato has rendered harmony a material object, one
akin to the lyre and its strings in their composite, physical, bodily,
earthy, and ‘mortal’ condition. In doing so he is showing himself to be
even more materialist and more reductionist than the later material-
istically inclined euphonists. For Plato, eager as he is to immortalize
the soul and to render all else ‘heavy, ponderous, [and] earthy’ (81c),
sound (harmony) cannot be allowed to be materially different from
its parts. The euphonists diverge from Plato in at least two respects,
but in both cases because they are operating with a cleavage within
sound that is unavailable to him: where he speaks of harmonia, they
speak of the σύνθεσις (combination, aggregate) of the letter-sounds
and the euphony that the letter-sounds give rise to. Plato’s gold
standard, by contrast, is the unrepeatable pure note that, to be arrived
at, must be segregated not only from other notes but also from the
conditions of its production.
First, for them sound is not clearly a composite entity. There is, as
we just saw, the σύνθεσις of the letter-sounds, and then there is the
euphony that ‘appears on its surface’ of the letter-sounds. A debate
raged amongst the κριτικοί as to which of these two entities ought to
provide the criterion of poetic excellence, but nowhere do we get the
slightest impression that euphony is just the synthesis of the elements
of individual sounds, as if one could tie all the sounds of a word or
verse into a neat sum and call this its euphony. (Another way of
underscoring this difference is to think of σύνθεσις as a verbal noun
rather than what linguists call a resultative noun: it represents the
combining of sounds, not their (blended) combination, which under
83
The premise is slipped in rather unexpectedly, then never relinquished in the
subsequent arguments.
228 James I. Porter

the best of conditions will be euphonious.) Rather, sound strikes the
ear as a collective φαντασία in a form that can no longer be analysed
into its original parts. Indeed, a well-made euphony will, precisely,
conceal its original parts, as we saw, and some of the most interesting
effects of euphony are no more than sound-effects that conceal the
fact that they have no original parts—they are a mere confection in
the ear of an auditor, a pure illusion of sound, like a false echo or a
referred pleasure or pleasurable pain.
Second, for the euphonists sound is and is not material. Here, the
euphonists can prevaricate in a way that Plato cannot because of the
aforementioned cleavage within sound that is available to them but
unavailable to him. Nowhere in their writings is it suggested that
euphony and σύνθεσις are materially reducible to each other, no
doubt because their intuition was to assume that euphony, being
literally ἐπιφαινομένη (‘appearing on the surface’), has the consistency
of an appearance; its materiality may therefore be different from
that of the material cause of the appearance, or it may not be,
depending upon one’s ontology. The euphonists don’t take a clear
stand on this issue, and so we can only make guesses on their behalf.
In terms that are closer to their own (seeing how expressions equiva-
lent to material are not quite part of their working vocabulary),
euphonious sound is and is not reducible to the constituent parts
(the sounds) that comprise it. At any rate, the euphonists are less
interested in pursuing ontologies than they are in pursuing the
mechanisms by which aesthetic sounds come to be produced or
released in prose and poetry.
The pursuit of the euphonists, too, is as much of the sound of
soundlessness as it is of sound. Recall that the final object of Crates’
critical analysis is not the sound of poems but the art (ὁ τῆς τέχης
λόγος) by which that sound is released.84 We might say that what
Crates is after is not sound but the structure of sound (εὔκρατος
ἁρμονία; Karin Pohl in her dissertation rightly speaks of ‘lautliche
Struktur’).85 And yet, given the desperateness of the search, the
elusiveness of the object, but also its quasi-materiality and the insist-
ence upon the absolute location of the structural properties of sound,
we might say that what Crates, at least half the time, is in fact after is
84
Phld. Poem. 5 col. 28 Mangoni; Porter (1995a) 99.
85
Pohl (1968) 152–3. Dionysius speaks of εὔκρατος ἁρμονία at Comp. 24,
120.11–12.

the ‘sound of the structure’ itself (what Webern, the composer, refers
to as the Strukturklang, as opposed to the Klangstruktur).86 Either
way, qua vocality or qua sound in relation, it is this ‘voice’ behind and
beyond the actual dimensions of vocality and sonority that is the
ultimate object of the euphonist critic’s quest. Their search is for the
sublimity of sound. As Longinus writes, ἁρμονία, composition in its
capacity to generate euphony, ‘echoes the sublimity’, ἡ ἁρμονία τῷ
ὕψει συνηχεῖ (39.4). It does not echo the sound; it echoes what lies
beyond sound, and resounds with it.87 The material sublime of sound
is, we might say, made out of a sublime material, namely one that
points us to this quality of voice that lies beyond sound itself.88
Longinus knows of Cicero’s sublimity: he adduces him in a σύγκ-
ρισις with Demosthenes and Plato: ‘Demosthenes has an abrupt
sublimity; Cicero spreads himself, . . . like a spreading conflagration’
(12.3–4); Plato’s ‘smooth style’ has a ‘soundless flow’ (ἀψοφητὶ ῥέων)
and is combined with grandeur (13.1). Does Longinus know of
Cicero’s sublime theory of ‘eloquence which rushes along with the
roar of a mighty spring, which all look up to and admire, and which
they despair of attaining’ (Or. 97)?89 Longinus’ recognition of the
rhetorical sublimity of Plato is in line, at least, with Cicero’s like
appraisal (13.4; Or. 62 and 67). There is a sly rhetoricization of
Plato here: in both authors Plato is being stood on his head.90
Which brings us back to our problem: what does Plato’s sublime
style, beyond the reaches of the voice (ἀψοφητί), sound like?
There is much more that could and should be said. In particular,
the parallels between Cicero, Longinus, and the κριτικοί could be
multiplied, and a deeper understanding of their various critical pro-
jects could be attempted.91 I have tried to suggest that euphonic
86
Žižek (1997) 47.
87
Cf. Longinus’ central thesis that ‘the sublime is an echo of a noble mind’ (9.2).
See Groddeck (1995), 70: ‘Das Erhabene ereignet sich zwischen Reden und Hören,
oder—anders gewendet—in dem seltsamen Raum zwischen Text und Lektüre.’
88
Similarly, language expressively delivered, with vocis mutationibus, reaches into
the resources of the body beyond the strictly linguistic: est enim actio quasi corporis
quaedam eloquentia (Or. 55).
89
Cf. Or. 39; 66; 67 (Kroll ad loc. compares Subl. 13.1).
90
See Or. 10, where ille non intellegendi solum sed etiam dicendi gravissimus auctor
et magister Plato has point; cf. also 51 and 63 on the style of the philosophers.
91
Cicero also knows, for instance, the thunderbolt metaphor as applied to Demos-
thenes (Or. 234; Subl. 34.4), with antecedents in Aristophanes (Ach. 530–1; Or. 29).
Cicero prefigures the agonistic (but also Longinian) commonplace that ambitious
230 James I. Porter

criticism points to a dimension beyond its apparent resources in
sound: those of us working on Philodemus have perhaps been talking
too much about sound! And if we take Dionysius as our cue, we have
been talking too much about euphony and not enough about disson-
ance (ἀντιτυπία), or about the interruption of consonance (ἀναβολή,
ἐγκοπὴ τῆς ἁρμονίας), of which hiatus is but one instance.92 I might
add that sublime criticism is also bound up with the critical and
ideological aspects of aesthetic criticism in antiquity, another aspect
we tend to neglect.93 Thus, a sensuous aesthetics needn’t be grounded
in naturalism, or even in the senses, even if it makes strategic use of
the postulates of naturalism to loosen the grip of deeply seated (and
naturalized) conventions of aesthetic perception and criticism. That is
one way of making sense of the difficulty of the critics I have been
discussing here: the paradox of soundless sound, the passage from
Klangstruktur to Strukturklang, is a provocation to think beyond
the limits of aesthetic discourse itself. Cicero’s polemics against his
Atticist opponents has a similar, pulling-up-the-rug-from-under-them
aspect to it; and his view has lasting if unsettling implications for
rhetorical theory as well.94
But a more immediate motivation for the theory these critics share
is to be found in the very situation of literary criticism. The focus on
the voice as a substance that is simultaneously an abstract and
spiritual entity (a spiritus or πνεῦμα) recalls the perpetual problem
of Greek literary culture: how to breathe life into the lifeless matter of
a text. Readers, Longinus writes, are ‘possessed by a spirit (πνεῦμα:
breath or voice) not their own’; sublimity makes ‘a kind of lustre
bloom upon our words as upon beautiful statues; it gives things life
(ψυχήν) and makes them speak (φωνητικήν)’ (13.2; 30.1). ‘Books,’
writers will risk greatness despite its impossibility as they try to rival the best writers
(Or. 3–4). And he recognizes sublimity as the true source of poetic marvel (grandior,
excelsior, magnificentius, fulmina, gravis acer ardens, etc.: Or. 15–16, 97–9, 119, etc.);
see Porter (2016) 280, 386, 612, for these and other parallels.
92
Comp. 20, 91.13–14; 22, 102.8; 103.5–6; 109.16–17; cf. Comp. 22, 104.5–6 (cit. n.
67 above).
93
See Porter (1994), 81; Porter (1995b), 136; Porter (2006).
94
Kroll (1913) 5 is partly right: ‘ein überaus geschickter literarischer Fechter’.
Despite ibid., 7 n. 1, Cicero invokes a preexisting sublime tradition that percolates
through the whole awareness of literary and rhetorical criticism and theory in
antiquity and that eventually crystallizes in works explicitly devoted to the topic, of
which those by Caecilius and Longinus have conventionally been the only recognized
instances. See Porter (2016).

Cicero writes, recalling a theme that, as we saw, occurs in Hierony-
mus of Rhodes, ‘lack that breath of life (spiritu illo) which usually
makes . . . passages seem more impressive when spoken than when
read’ (Or. 130).95 And it is probably the euphonist Pausimachus
who says, ‘But when the <verses> of Homer are read out (ἀναγινώσ|
κητ]α̣ι̣), all <the events> appear greater and more beautiful (πάντα
μ[ε]ίζω | καὶ κα]λ̣λίω φ[αίνε]ται). . . . ’96 But how does a written text
sound? The κριτικοί and Longinus offer us, among other things, a
theory of reading, a way of reading the voice buried in the voiceless
script of Greek texts from the distant past. The written text has a
sound that cannot be heard.97 Bringing out its hidden music while
respecting its precious concealment is the difficult task of sublime
criticism.
95
See Porter (2010) 330–40 on ἔμψυχος. The related term πνεῦμα, used in the sense
of the breath of the musical voice, is found in Dionysius of Halicarnassus, e.g. Comp.
14. Aristoxenus may be the proximate origin of the term (Kroll (1907) 97), but he was
by no means the first to πνεῦμα in this sense.
96
Poem. 1 col. 43.9–12.
97
Cf. Dion. Hal. Dem. 22, 177.21–178.2: ‘If, then, the spirit (πνεῦμα) with which
Demosthenes’ pages (τοῖς βιβλίοις) are still imbued after so many years possesses so
much power (ἰσχύν) and moves his readers in this way (ἀγωγόν), surely to hear him
delivering his speeches at the time must have been an extraordinary and overwhelm-
ing experience (ὑπερφυές τι καὶ δεινὸν χρῆμα).’ The epithets are all Longinian. In
Comp. 22, 99–100 Dionysius preserves Pindar fr. 75 S-M, which prominently the-
matizes the voice as echo (cf. ibid., 11, 42.1–3, treating the sound of silence). Cf. also
Dio Or. 36.27, where Dio’s interlocutors, the rude Borysthenians ask him to approxi-
mate Plato’s ‘nobility of expression (φράσις)’: ‘for if we understand nothing else, we do
understand at least his language because of our long familiarity with it, for it has a
lofty sound (οὐ σμικρόν), not far removed from the voice of Homer (τῆς γε φωνῆς . . .
οὐδὲ πόρρω τοῦ Ὁμήρου)’.
10
Disreputable Music
A Performance, a Defence, and their
Intertextual and Intermedial Resonances
(Plutarch Quaest. conv. 704c4–705b6)
Andrew Barker
Even if we read it ‘straight’, without burrowing into its intertextual

cellarage, the beginning of Quaest. conv. 7.5 (704c4–705b6) is
remarkable enough. It describes a performance of music of which
Plutarch heartily disapproves, and continues with a speech arguing in
defence of music of that sort. After this comes a second and much
longer speech (705b7–706e11) in which this disreputable kind of
music is criticized on moral grounds, most of which are fairly famil-
iar; but that speech falls outside the scope of the present chapter.
References to unseemly music are common enough in ancient litera-
ture, but performances of it are seldom described in such vivid detail,
and reasoned defences of such music are even rarer. But the passage
becomes even more interesting if we explore its underground terri-
tory, that is, its implicit allusions to well-known earlier writings, and
try to understand the relation between the basement and the upper
storeys. Most of this chapter is an attempt to follow some of the
indications in the text which give access to this hidden substratum;
and I shall offer some thoughts about the ways in which they encour-
age us to interpret what Plutarch has actually written.1
1
On the value of intertextuality in interpreting the Quaest. conv., see Ruffy (2012)
12–15.
234 Andrew Barker

But the chapter has a second purpose too, hinted at by the word
‘intermedial’ in its title. This text is of course not designed to be sung
or musically accompanied; unlike some kinds of poetry, its meaning
is not modified by being expressed through musical performance.
On the other hand, it is the account of a musical performance that
sets the scene for the subsequent discussion. Although the explicit
focus of the speeches that follow is on the merits and defects of certain
types of music rather than on that particular performance, what
provides the occasion for the discussion is the live performance itself;
the speeches take their cue from the performance, and from the
emotionally charged ways in which it affected its hearers (including,
of course, the speakers themselves). This holds especially of the speech
we shall examine here, not just because it immediately follows the
performance, but more importantly because it is this performance in
particular that the speaker is principally trying to defend. He is anxious
to disarm criticism of the music the symposiasts have just heard (for
whose presentation he is partly responsible), and of their reactions to it.
Now plainly the dialogue’s readers do not hear the music. I want to
suggest, however that they become, as it were, a ‘virtual audience’ of
the performance Plutarch has described. His colourful depiction of it
presents them with a scenario which their own experiences enable
them to recreate in their imaginations, and thus invites them to
respond to the speeches in the light of their own past experience of
music as well as of texts. Echoes of the performance, as their experi-
ences have led them to envisage it, resonating in their aural imagin-
ation while they read the speeches, will channel influences through a
non-textual medium into their reception of what they read. This
input from musical experience into the interpretation of the text is
not, perhaps, an instance of precisely the kind of ‘intermediality’
discussed in the Introduction to this volume, but the relation between
them is evidently close; in both cases something conveyed through
one cognitive or perceptual medium affects our responses to some-
thing conveyed through another. But however plausible this sugges-
tion may be, it is of no great interest unless we can apply it in
illuminating ways to the study of the text we are scrutinizing. Can
we in fact do so? If so how, and how, if at all, do influences of this sort
interact with those of the recurrent intertextual allusions? That is the
second group of issues which this chapter will try to address, but the
passage’s intertextual intricacies need to be identified and examined
in their own right before we consider them.
Disreputable Music 235

INTERTEXTUAL ALLUSIONS
(a) The Performance
ἐν Πυθίοις Καλλίστρατος, τῶν Ἀμφικτυόνων ἐπιμελητής, αὐλῳδόν

τινα πολίτην καὶ φίλον ὑστερήσαντα τῆς ἀπογραφῆς τοῦ μὲν
ἀγῶνος εἶρξε κατὰ τὸν νόμον, ἑστιῶν δ’ ἡμᾶς παρήγαγεν εἰς τὸ
συμπόσιον ἐσθῆτι καὶ στεφάνοις, ὥσπερ ἐν ἀγῶνι, μετὰ τοῦ χοροῦ
κεκοσμημένον ἐκπρεπῶς. καὶ νὴ Δία κομψὸν ἦν ἀκρόαμα τὸ
πρῶτον· ἔπειτα διασείσας καὶ διακωδωνίσας τὸ συμπόσιον,
ὡς ᾐσθάνετο τοὺς πολλοὺς ἐγκεκλικότας καὶ παρέχοντας ὑφ’
ἡδονῆς ὅ τι βούλοιτο χρῆσθαι καὶ καταυλεῖν καὶ ἀκολασταίνειν,
ἀποκαλυψάμενος παντάπασιν ἐπεδείξατο τὴν μουσικὴν παντὸς
οἴνου μᾶλλον μεθύσκουσαν τοὺς ὅπως ἔτυχεν καὶ ἀνέδην αὐτῆς
ἐμφορουμένους· οὐδὲ γὰρ κατακειμένοις ἔτι βοᾶν ἐξήρκει καὶ
κροτεῖν, ἀλλὰ τελευτῶντες ἀνεπήδων οἱ πολλοὶ καὶ συνεκινοῦντο
κινήσεις ἀνελευθέρους, πρεπούσας δὲ τοῖς κρούμασιν ἐκείνοις καὶ
τοῖς μέλεσιν.
[704c4] In the course of the Pythia, Callistratus, the president of
the Amphyctyons, prevented an aulode—a citizen and a
friend—from registering for the contest, in compliance with
the regulations, because he was late. But when he entertained
us to dinner, he brought him to the symposium, splendidly
decked out, along with the χορός, in the costume and wreath
he would have worn in a contest. And what we heard at first was
indeed most impressive. [d] But then, when he’d taken the
measure of the symposium and sounded it out, seeing that
most of those present had yielded to pleasure and would let
him play whatever he liked, casting spells on them with his
piping and indulging his licentious tastes, he completely
unmasked himself [or ‘stripped off ’] and performed music far
more capable than any wine of intoxicating people who drink it
up carelessly and without restraint. They weren’t even satisfied
with chanting and clapping while they reclined; in the end most
of them leaped up and danced together with movements quite
unfit for free men, though perfectly suited to those instrumental
sounds and tunes.
The musician is not a guest of the normal sort. The party seems to be
in full swing when he arrives, wreathed and in his musician’s finery,
along with a χορός (whose identity poses problems to which I’ll return
immediately), and the first part of his performance, we are told, was
236 Andrew Barker

most impressive, a κομψὸν ἀκρόαμα (704c9–d1); later, however, his
presence becomes thoroughly disruptive. In a general way, this is a
very familiar theme. I have no doubt at all that Plutarch designed it as
a deliberate and none too subtle reminder of Alcibiades’ irruption
into the party in Plato’s Symposium; and as we’ll see in due course, the
Symposium, especially but not only the Alcibiades passage, is the most
important of the texts underlying the description that follows.2
The reference to a χορός, however, is a little puzzling. We would
expect a musician’s χορός to be a group of singers or dancers who
took part in the performances in which he took the starring role, and
would have done so at Delphi if they had not been disqualified. But
Plutarch gives no indication that the disqualification affected a choral
performance; on the contrary, it was apparently just this one musi-
cian who was due to register but failed to do so on time, and who was
therefore prevented from competing, presumably as a soloist. Again,
if this χορός is a group of performers, there would seem to be no point
in mentioning them unless they are actually involved in the music-
making at Callistratus’ symposium; yet as far as we can tell from
Plutarch’s account, they are not. It is therefore tempting to suppose
that the word χορός is being used here in a non-musical sense, as it
not infrequently is, in which it refers to a prominent individual’s
entourage, his band of followers and supporters. To suppose that it
has this meaning here may strain our credulity, but perhaps Plutarch
did indeed intend readers to construe it in that sense, while enjoying
the ambiguity created by its presence in this largely musical context
(I shall return to this point in Part 2).
For the present I shall leave the question of the identity of the χορός
unresolved. But given that it has no musical role in the scenario, we
may wonder whether there is any detectable reason—apart from the
general context of musical activity—for Plutarch’s choice of this
word. Perhaps there is. Alcibiades’ retinue in the Symposium is not
called a χορός, but that of a star performer in another Platonic
dialogue is indeed so described; it is that of the sophist Protagoras,
mentioned in the course of the narrative introducing the dialogue
2
Plutarch introduces his preface to the Quaest. conv. with an allusion to the
Symposium (at 613d), and it seems to have been much in his mind throughout its
composition: as Frieda Klotz remarks, ‘the Symposium is a crucial model for the
Table Talk’ (Klotz (2014) 212; her essay can be recommended as a thought-provoking
study of the work as a whole).

named after him (Prt. 315b). And what the sophists depicted in that
passage produce—especially in the booming and unintelligible utter-
ances of Prodicus—could very aptly be described as a κομψὸν ἀκρόαμα.
Further, one of the key features of sophists in Plato’s dialogues, espe-
cially Protagoras and Gorgias, is that their speeches are (in Plutarch’s
words) a κομψὸν ἀκρόαμα τὸ πρῶτον but only τὸ πρῶτον: when critic-
ally assessed they are worthless and potentially pernicious.3 Exactly the
same is true of the performance of Plutarch’s musician, as we shall see.
We might usefully remember that there is a strong link between this
part of the Protagoras and the Symposium; with the exception of
Aristophanes, all the main characters of the Symposium are there in
the Protagoras, listening to the sophists in Callias’ house. (Apart from
Socrates himself, Aristophanes is the only one of them who is not
implicitly tarred with the brush of sophistry either here or in the
Symposium, or indeed anywhere else in the dialogues.) If I am right
in construing Plutarch’s words as pointers to the Protagoras, they will
suggest that there is a marked affinity between the musician’s perform-
ance and the dangerous teachings of the sophists. The superficiality of
his charms may also be suggested by the allusion to his magnificent
costume; we can find a similar use of this motif in an oration by the
Platonizing emperor Julian, for instance, where he says that the people
who will be impressed by a kitharode’s costume and the appearance of
his splendid instrument are just children, or men and women with the
mentality of children (Or. 3.7).
The allusions to this part of the Protagoras direct us back to the
Symposium, and to Plutarch’s implicit references to it. He drops a
broad hint even before describing the musician’s dramatic entrance;
he had been disqualified from competing in the Pythia because he was
late in registering for the contest, ὑστερήσαντα (704c5).4 Late arrival
is a minor motif in the Symposium; Socrates is late, having been
detained by a communication from his daimonion, and Alcibiades
only arrives when the evening is well advanced. And whereas
3
The adjective κομψός and its cognates are common in Plato; not always, but very
often, they are used ironically, to suggest that the things so described look or sound
impressive at first sight but are in fact sophistic or absurd. Examples in Rep. include
405d, 436d, 460a, 489b, 499a (where it is coupled with ἐριστικά and unfounded δόξα),
572c. See. e.g. Gorg. 521e, referring back sarcastically to Callicles’ usage at 486c; Crat.
426a; Phd. 101c; Lach. 197d; Tht. 171a; Phaedr. 230c.
4
The rules about late registration were probably as strict in the Pythia as they seem
to have been at Olympia, on which see Poliakoff (1987) 19–20.
238 Andrew Barker

Socrates’ influence on the company is wholly edifying, Alcibiades’
uninvited invasion (Smp. 212d–213a) immediately disrupts the sobri-
ety of the gathering; and towards the end of the dialogue, after
another late invasion by uninvited party-goers, the symposium des-
cends into a bout of uninhibited drinking and revelling (223b). In
much the same way the musician’s performance, evidently not right
at the beginning of the party, causes the complete breakdown of
civilized decorum among the guests in the house of Callistratus.
I shall say no more about the description of the musician’s entrance
and the admirable music which he presents τὸ πρῶτον. While doing
this, it turns out, he has been watching the audience’s reactions and
sizing them up, as it were, διασείσας καὶ διακωδωνίσας τὸ συμπόσιον;5
and we might be reminded of Ion the rhapsode keeping a calculating
eye on his audience while expressing the extremities of Homeric
passion and grief. If they’re weeping he’ll be laughing all the way to
the bank, and if they’re laughing he may find himself in serious
financial trouble (Plato, Ion 535e). But perhaps there’s nothing to
be made of that. The verb διακωδωνίζειν is uncommon enough to
attract attention,6 but there is nothing in its earlier occurrences, as far
as I can see, that would be relevant to the present discussion. By
contrast, the innocent-looking expression τὸ συμπόσιον might be
significant. Of course in one sense it is perfectly appropriate—after
all, this is a symposium—but its precise usage here, as a collective
noun for the participants at a symposium, strikes me as unusual and
5
The translation of this phrase in the Loeb edition, ‘shaking the hall and filling it
with resounding noise’, not only gives τὸ συμπόσιον a most improbable sense, but also
misses the connection of the two verbs with processes of testing and assessing.
διασείειν is used in the sense ‘to sift’, as in ‘sifting the wheat from the chaff ’.
διακωδωνίζειν is commonly used of making a coin ‘ring’ to check whether it is
genuine, and there are some intriguing explanations of the origin of this and similar
usages. A scholium on Ar. Birds 842 explains that people went round the guard-posts
at night and rang a bell (κώδων), which the guards had to answer to show that they
were awake (perhaps, the scholiast adds, Aristophanes’ allusion is a parody of an
episode in Euripides’ Palamedes). This explanation reappears at Etym. magn. 273.47,
which interprets διακωδωνίζειν as ‘to test and assess’, and offers two other possible
explanations, deriving it either from the practice of ringing a bell to test the mettle of
fighting quails, or from the use of the same technique to assess the ‘nobility’ or courage
of horses, by seeing whether or not it frightens them.
6
The TLG records some ninety occurrences in Greek literature, but the great
majority are from very late sources.

rather odd.7 We would surely have expected something like τοὺς
ἀκουόντας or τοὺς παρόντας, and the substitution of τὸ συμπόσιον
invites our curiosity. Could we possibly interpret it as Plutarch’s way
of waking up his less perceptive readers to the fact that it’s ‘the’
Symposium, Plato’s Symposium, that should be in their minds? It is
impossible to be sure, but I suggest that it’s a colourable hypothesis.
The musician’s assessment of his audience shows him that they are
putty in his hands, and that in their pleasure at his music they will let
him do whatever he likes, καὶ καταυλεῖν καὶ ἀκολασταίνειν, casting
spells on them with his piping (καταυλεῖν) and indulging in unre-
strained licentiousness (704d2–4). Here we must stray away from the
Symposium again for a moment; καταυλεῖν occurs only three times in
Plato, twice in the Republic and once in the Laws, and it’s the
Republic’s usages that are significant here. At 411a we find the verb
in Socrates’ depiction of a person who risks doing serious damage to
himself ‘when he allows (παρέχῃ, as in Plutarch) music καταυλεῖν καὶ
καταχεῖν τῆς ψυχῆς by pouring sweet and soft and mournful tunes
through his ears as if through a funnel’.8 Later, at 561c–d, Socrates
catalogues the diverse recreations of the ‘democratic’ individual, who
makes no distinction between good and bad pleasures, and at the top
of the list are episodes in which he is μεθύων καὶ καταυλούμενος,
fuddled with drink and spellbound by the piping of the aulos. Thus
even without the explicitly pejorative connotations of Plutarch’s
ἀκολασταίνειν, Plato’s treatment of the verb καταυλεῖν would be
enough to sound a note of warning. The fact that the audience’s
pleasure will allow the musician to do whatever he likes corresponds
exactly to the ‘democratic’ man’s indiscriminate openness to pleas-
ures of any sort, which allows him to wallow in any and all of them.
The next clause unmistakably takes us back to the Symposium,
in what is perhaps the most important allusion of them all:
ἀποκαλυψάμενος παντάπασιν ἐπεδείξατο τὴν μουσικὴν παντὸς οἴνου
μᾶλλον μεθύσκουσαν τοὺς ὅπως ἔτυχεν καὶ ἀνέδην αὐτῆς ἐμφορουμένους
(704d4–6). The musician metaphorically strips himself naked and
reveals what he’s really up to, with a display of the kind of music that
is more inebriating even than wine, for those who swallow it without
due caution. In the Symposium Alcibiades famously likens Socrates
7
Teodorsson (1996) 66 identifies parallel uses of the word in Plutarch at Quaest.
conv. 710c and Sept. sap. 157d, and rather less plausibly at Sept. sap. 164d.
8
On this passage, see further Peponi (this volume) pp. 172–4.
240 Andrew Barker

to an aulete, specifically the aulete Marsyas, and starting with his
image of the statuettes of silēnoi nested inside one another like
Russian dolls (Smp. 215a–b), he describes the process of uncovering
the treasures he contains by peeling off layer after layer. Later, in his
vivid account of his encounters with Socrates, both he and Socrates
are both metaphorically and literally stripped naked. They strip
down for wrestling, and Alcibiades snuggles up to Socrates in bed,
but in neither case does he get the reaction he expected; the Socrates
he unveils is solid virtue and wisdom right to the core (217b–219d).
What Plutarch presents when the musician abandons his respectable
guise is an inversion of Plato’s revelations about Socrates; it is more
like a ‘candid camera’ shot of Alcibiades in one of the more shame-
lessly unregenerate phases of his chaotic career.
I suggest that this is the key to the whole description. It portrays a
set of actions and reactions in which everything that Plato stands for
in the sphere of music and ethics is turned upside down, and the
allusions to the dialogues tell us that that is what it is. Resonances of
other passages in Plato can also be found without difficulty, parallels
in the Republic, for instance, of Plutarch’s evocation of the κινήσεις
ἀνελευθέρους to which the music impels the symposiasts at 704d6–10
(notably Rep. 400b, cf. 395c, 401b). Disorderly music and indiscrim-
inate drinking are rejected in the preliminaries to the Symposium
(176a–e), and their proper control and conjunction is a major topic in
the Laws. One might consider also the various Platonic depictions
and discussions of μανία, from which drinking (rather curiously)
eventually rescues the company at 704e1, and other passages too
might justifiably be cited. But I shall leave the musician’s perform-
ance there.
(b) The Defence Speech
ἐπεὶ δ’ ἐπαύσαντο καὶ κατάστασιν αὖθις [704E] ὥσπερ ἐκ μανίας

ὁ πότος ἐλάμβανεν, ἐβούλετο μὲν ὁ Λαμπρίας εἰπεῖν τι καὶ
παρρησιάσασθαι πρὸς τοὺς νέους· ὀρρωδοῦντι δ’ ὅμως αὐτῷ μὴ
λίαν ἀηδὴς γένηται καὶ λυπηρός, αὐτὸς ὁ Καλλίστρατος ὥσπερ
ἐνδόσιμον παρέσχε τοιαῦτά τινα διαλεχθείς· ‘Ἀκρασίας μέν’, ἔφη,
‘καὶ αὐτὸς ἀπολύω τὸ φιλήκοον καὶ φιλοθέαμον· οὐ μὴν Ἀριστοξένῳ
γε συμφέρομαι παντάπασι, ταύταις μόναις φάσκοντι ταῖς ἡδοναῖς τὸ
‘καλῶς’ ἐπιλέγεσθαι· καὶ γὰρ ὄψα καλὰ καὶ μύρα καλοῦσι καὶ καλῶς
γεγονέναι λέγουσιν δειπνήσαντες ἡδέως καὶ πολυτελῶς. δοκεῖ δέ μοι

μηδ’ Ἀριστοτέλης αἰτίᾳ [F] δικαίᾳ τὰς περὶ θέαν καὶ ἀκρόασιν
εὐπαθείας ἀπολύειν ἀκρασίας, ὡς μόνας ἀνθρωπικὰς οὔσας, ταῖς δ’
ἄλλαις καὶ τὰ θηρία φύσιν ἔχοντα χρῆσθαι καὶ κοινωνεῖν. ὁρῶμεν
γὰρ ὅτι καὶ μουσικῇ πολλὰ κηλεῖται τῶν ἀλόγων, ὥσπερ ἔλαφοι
σύριγξιν, ἵπποις δὲ μιγνυμέναις ἐπαυλεῖται νόμος, ὃν ἱππόθορον
ὀνομάζουσιν· ὁ δὲ Πίνδαρός φησι κεκινῆσθαι πρὸς ᾠδήν
ἁλίου δελφῖνος ὑπόκρισιν·
[705A] τὸν μὲν ἀκύμονος ἐν πόντου πελάγει
αὐλῶν ἐκίνησ’ ἐρατὸν μέλος·
ὀρχούμενοι δὲ τοὺς ὤτους αἱροῦσι, χαίροντας τῇ ὄψει καὶ μιμητι-
κῶς ἅμα δεῦρο κἀκεῖσε τοὺς ὤμους συνδιαφέροντας. οὐδὲν οὖν
ὁρῶ τὰς τοιαύτας ἡδονὰς ἴδιον ἐχούσας, <ἢ> ὅτι μόναι τῆς ψυχῆς
εἰσιν, αἱ δ’ ἄλλαι τοῦ σώματος καὶ περὶ τὸ σῶμα καταλήγουσιν·
μέλος δὲ καὶ ῥυθμὸς καὶ ὄρχησις καὶ ᾠδὴ παραμειψάμεναι τὴν
αἴσθησιν ἐν τῷ χαίροντι τῆς ψυχῆς ἀπερείδονται τὸ ἐπιτερπὲς
καὶ γαργαλίζον. ὅθεν οὐδεμία τῶν τοιούτων ἡδονῶν ἀπόκρυφός
ἐστιν οὐδὲ σκότους δεομένη καὶ τῶν τοίχων ‘περιθεόντων’,
ὡς [β]οἱ Κυρηναϊκοὶ λέγουσιν, ἀλλὰ καὶ στάδια ταύταις καὶ
θέατρα ποιεῖται, καὶ τὸ μετὰ πολλῶν θεάσασθαί τι καὶ ἀκοῦσαι
ἐπιτερπέστερόν ἐστι καὶ σεμνότερον, οὐκ ἀκρασίας δήπου καὶ
ἡδυπαθείας ἀλλ’ ἐλευθερίου διατριβῆς καὶ ἀστείας μάρτυρας
ἡμῶν ὅτι πλείστους λαμβανόντων.
When they had finished, and the drink had brought them back
to calm, [e] as if from a fit of madness, Lamprias wanted to say
something and speak his mind to the young men; but since he
was worried that he would be too stern and harsh, Callistratus
himself provided a prelude, as it were, in roughly the following
words.
‘I too believe that the love of things heard and seen is exempt
from akrasia; but I don’t altogether agree with Aristoxenus,
when he asserts that it is only these pleasures that are called
kalai. For people call foods and perfumes kala, and say that
things have gone kalōs when they have dined enjoyably and
sumptuously. Nor do I think that Aristotle exempted enjoyment
of things seen and heard from akrasia [f] for the right reason,
that is, on the grounds that they are specific to humans, whereas
creatures with the nature of wild beasts also experience and
share in the others. For we see that music also enchants many
of the irrational animals, as for instance deer are enchanted by
Panpipes; and people play the pipe-tune they call the hippothoros
nomos to mating mares. And Pindar says that “the dolphin of the
sea answers” in response to song, the dolphin [705a] “which the
242 Andrew Barker

lovely melody of auloi arouses in the wave-less open sea.” People
catch owls by dancing, as they delight in the sight and imitate it by
moving their shoulders back and forth in time with it. So I don’t
see anything peculiar in pleasures of these sorts, except that they
are pleasures of the soul, while the others are pleasures of the body
and come to an end in the body. But melody and rhythm and
dance and song travel past sense-perception and deposit delightful
titillation in the part of the soul that experiences pleasure. Hence
no pleasure of this sort is secret and needs the cover of darkness
and of walls ‘running around’ it, as [b] the Cyrenaics say. On the
contrary, stadia and theatres are built for them, and watching and
listening to something in company with many people is more
enjoyable and honourable; and thus we have a great many wit-
nesses to the fact that it is not akrasia and self-indulgence, but a
pastime fit for people who are free and civilized.’
When the performance is over and everyone has settled down, Lam-
prias feels like giving the young folk a piece of his mind, but hesitates to
cast himself as a carping curmudgeon, and Callistratus comes to the
rescue with a speech which, we are told, provides ‘as it were an
ἐνδόσιμον’ (704e1–5). ἐνδόσιμον is an interesting word, the nominal
form of an adjective commonly connected with music. It can refer to a
prelude to a song, or sometimes (as in Aristotle at Rhet. 1414b24 and
1415a8) to something like the ‘key-note’ of a melody; and these musical
uses are often deployed metaphorically, to refer to some other kind of
introduction or starting-point. But here, clearly enough, Callistratus’
speech in defence of allegedly disreputable music can hardly be the key-
note for Lamprias’ subsequent speech for the prosecution, and perhaps
we should not assign it any musical significance. What it does is to give
Lamprias a ‘pretext’ for his intervention, by supplying him with an
argument to which he can respond.9 In this way his denunciation of
such music will not come across as a direct reproof to the symposiasts;
instead it will follow the civilized conventions of a decorous Plutarchan
symposium, in which a second speaker regularly provides a counter-
argument to a thesis propounded by the first.10
9
This seems to correspond to Plutarch’s regular use of the word (which appears
in his writings eight times in the singular and once in the plural); an ἐνδόσιμον is
something that provides a pretext or ‘way in’ for some further action.
10
Roskam (2009) 377 describes Callistratus’ intervention as ‘adding fuel to the
flames’, which strikes me as missing the point.

Something rather remarkable happens at this point in the passage,
precisely at the moment when we move from the description of the
performance to the speech in its defence. So far the text has been
studded with allusions to Plato’s dialogues, but now we leave the
Academy behind and enter the world of the Lyceum. Callistratus
marks the transition right at the beginning of his speech, and he
does so in two ways. First, he associates himself with people who deny
that the love of things heard or things seen can lead to the vice of
ἀκρασία, that is, to an inability to resist the temptation of damaging
over-indulgence in such experiences (‘ἀκρασίας μέν’, ἔφη, ‘καὶ αὐτὸς
ἀπολύω τὸ φιλήκοον καὶ φιλοθέαμον’, 704e6–7). His thesis, essentially,
is that giving free rein to one’s love of sights and sounds can do no
harm, and that in this context the notion of over-indulgence simply
does not apply. But of course the most famous φιλοθεάμονες and
φιλήκοοι in Greek literature are the ‘lovers of sights and sounds’
who are contrasted unfavourably with philosophers towards the end
of Book V of Plato’s Republic (475d–476b). Socrates does not accuse
them of ἀκρασία (a noun that appears nowhere in Plato), but by
implicitly associating himself with them Callistratus clearly puts a
substantial distance between himself and Plato’s followers.
At the same time, secondly, he signals the transition to a Peripatetic
milieu, an environment to which Plutarch’s attitude seems at best
ambivalent.11 ‘I too (καὶ αὐτός),’ he says, in the sentence quoted
above, ‘believe that the love of things heard and seen is exempt
from ἀκρασία’; and it soon becomes clear that when he says ‘I too’,
the others with whom he shares this view are Peripatetics. He asso-
ciates himself with them at first in a rather back-handed way, by
disagreeing explicitly with two of their most eminent representatives,
Aristoxenus and Aristotle himself. Aristoxenus had apparently
claimed that only visible and audible things are described as καλά;12
Callistratus rejects this thesis and offers phrases referring to food and
perfume to prove his point (704e7–11). No doubt Aristoxenus was
aware of such usages, but would have treated them as vulgar
11
For a useful but perhaps over-cautious survey of Plutarch’s treatment of Aristotle
see Sandbach (1982) and on Quaest. conv. in particular Oikonomopoulou (2011).
Lopes (2009) 419 points out that Plutarch’s frequent citation of Aristotelian scientific
doctrines in Quaest. conv. does not show that he accepted them, and adds: ‘for
the most part these “quotations” are used either to get a discussion started . . . or,
less frequently, they are simply refuted’.
12
Aristox. fr. 74 Wehrli; cf. fr. 73.
244 Andrew Barker

solecisms; his real contention, I suppose, was that the adjective can
only properly be applied to things that we see or hear.13 But Calli-
stratus does not explain how his view on this issue is relevant to his
claim about ἀκρασία. Its only detectable effect is to assimilate lavish
feasting and luxuriating in delicious scents to the same evaluative
category as the pleasures of sights and sounds, since all, equally, can
be regarded as κάλα. Does he mean to imply that they too are exempt
from accusations of ἀκρασία? The implication would hardly pass
muster with a Platonist, or (we may suppose) with respectable citizens
of Delphi assembled at Callistratus’ symposium.
Callistratus’ disagreement with Aristotle is more qualified and
more interesting; he agrees with his conclusion—which is indeed
essential to his position—but not with the reason he gives for it
(704e11–f3). Here the obvious references are to two parallel passages
on σωφροσύνη in the Ethics (EN 3.10, 1117b–1118b, EE 1230b–
1231a). The conclusion, as Callistratus expresses it, is that indulgence
in the pleasures of sight and hearing cannot amount to ἀκρασία;
Aristotle’s reason, which Callistratus disputes, is that these pleasures
are the only ones peculiar to human beings, whereas those in which
overindulgence can lead us astray, betraying our human nature, are
those which we share with other animals.
Aristotle does indeed say something like this, but Callistratus is
playing fast and loose with him none the less. His own explanation of
the distinction between pleasures that can involve us in ἀκρασία and
those that cannot is that the former are purely bodily while the latter
penetrate to the soul; and he asserts that the pleasures of sight and
hearing belong to the latter class (705a3–8). Here too he seems to be
crossing swords with Aristotle, since this passage of the Nicomachean
Ethics treats the pleasures of sight and hearing as bodily (1118a1–9).
But elsewhere in Aristotle’s discussions of what we see and hear,
implicitly in other passages of the Ethics and very overtly in his
study of music in Book 8 of the Politics, it is perfectly clear that
these pleasures also have effects on the soul; and perhaps the apparent
contradiction with the passages on σωφροσύνη in the Ethics can be
13
He is not alone in taking this view. A short time after Plutarch, Ptolemy
champions the same thesis, though for emphatically un-Aristoxenian reasons, in a
fascinating passage of his Harmonics (3.3, 93.11–94.1 Düring); and it has some
affinities with Aristotle’s assertion in Book VIII of the Politics that among the objects
of sense-perception, only those that are audible possess ἦθος (Pol. 1340a28–b10; cf.
[Arist.] Pr. 19.27). But of course that is not quite the same proposition.

resolved, though I’m not sure how that would be done. We may also
note that in Callistratus’ version the issue is to do with ἀκρασία,
whereas Aristotle does not mention ἀκρασία in these passages; the
failing to which he makes the relevant allusions is ἀκολασία.14 But
perhaps in the present context that is not very important.15
What is more interesting is that the Aristotelian ingredients of the
defence speech don’t end with these explicit references. Aristotle
asserts in both versions of the Ethics that animals other than humans
do not take pleasure in things they see or hear as such.16 They respond
to them only as indications, for instance, of a potential meal; and the
Eudemian Ethics explicitly denies that animals are capable of perceiv-
ing εὐαρμοστία, the quality that makes sounds musical, or κάλλος,
beauty in general (1230b36–1231a4). Callistratus devotes the next
nine lines of his speech to a series of examples designed to show
that this is untrue, and that animals do derive pleasure directly from
their visual and auditory experiences (704f3–705a3).17 Deer, horses,
and dolphins respond enthusiastically to music, he says, and owls
delight in the sight of dancing, waggling their shoulders back and
forth in imitation (μιμητικῶς) of what they see—a phenomenon
apparently exploited by the Greek counterparts of Papageno.18 My
point is that two of the four examples are taken straight from
Aristotle. I have found nothing relevant about the dolphin in his
writings, or about horses and the ἱππόθορος νόμος to which Callistratus
refers (704f6),19 but the others originate in the Historia animalium
14
This point is noted by Sandbach (1982) 220. He concludes that Plutarch may not
have had direct access to the text of the Ethics, but his grounds for this conclusion are
thin. He notes also that ἀκρασία appears in relevant contexts at [Arist.] Pr. 28.2, 3, and 7.
15
In substituting ἀκρασία for Aristotle’s ἀκολαστία Callistratus might be exercising
a little social tact. Acquitting them of ἀκρασία suggests that the symposiasts have high
moral standards, and reassures them that in responding to the music as they did, they
were not abandoning them. Acquitting them of ἀκολαστία, by contrast, would be
relevant only if they might plausibly be suspected, however unjustly, of lacking such
standards altogether, and being prone to unbridled licentiousness.
16
They derive pleasure directly only from experiences of touch and taste. Aris-
totle’s contentions are repeated, with minor variations, at [Arist.] Pr. 28.7.
17
Plutarch may have had some sympathy with this position. It would chime with
his attribution to animals of some degree of reason and moral sensitivity, on which see
Newmyer (2014) 226–31.
18
For other references to these phenomena see Teodorsson (1996) ad loc.
19
Plutarch mentions this nomos again at Coniugalia praecepta 138b; I have not
been able to identify the source of his information. Clement of Alexandria, in his
allusion to the practice of using aulos music to encourage mares that are being mated,
remarks that the μουσικοί call this music ἱππόθορος, and the fact that he refers to οἱ
246 Andrew Barker

(HA 597b21–9, 611b26–31);20 and the information that Aristotle
provides would be enough by itself to support this part of Callistratus’
argument. We might also note that in Book 8 of the Politics he seems
to deny what he had said in the Ethics, asserting that music of a
popular sort is enjoyed ‘even by some of the animals’, as well as by
‘the common run of slaves and children’ (Pol. 1341a15–17).
It appears, then, that Callistratus has constructed his argument by
exploiting inconsistencies between passages in Aristotle’s writings,
using Aristotle to confute Aristotle, and I’m not sure what we should
make of that. Here again, indeed, as in the comments on Aristoxenus,
it’s not immediately clear how this Peripatetic excursion is supposed
to help in establishing the thesis that Callistratus is trying to sustain,
that is, that the pleasures of sight and hearing cannot lure us into
ἀκρασία. This is a thesis of whose truth he must evidently try to
convince us, but it seems to make no difference to his argument
whether the reason for it is the one given by Aristotle or his own,
which is that only the pleasures of sight and hearing ‘are pleasures of
the soul, while the others are pleasures of the body and come to an
end in the body’ (705a5–6). Or, if it does make a difference, it must be
because his explanation embraces the view that these pleasures affect
the soul as well as the body. But if that is the point, it would appear to
give a valuable hostage to fortune, since it offers a Platonist just the
opening he needs to launch on his counterarguments. This might just
possibly be what Plutarch meant when he described Callistratus as
providing Lysias with an ἐνδόσιμον. If it were, ‘key-note’ would be an
apt translation after all; Callistratus’ contention is a point on which
both parties agree, and on which Lamprias can safely rely in his
rebuttal.
Even so, Callistratus could have used his thesis to his advantage,
and perhaps he implicitly does. He insists, as he must, that auditory
and visual pleasures are exempt from ἀκρασία, but so far he has
offered no arguments in support of that position. But if we elide, as
he does, the difference between ἀκρασία and ἀκολασία, or assume that
μουσικοί rather than to people in general suggests that it was a term coined by musical
specialists. Perhaps it was used only by them, and was not current in ordinary usage or
among horse-breeders themselves.
20
Sandbach (1982) 225–7 is devoted to showing that Plutarch knew the HA, citing
e.g. De sol. an. 973a, 979c–e, 981f. He presents inter alia a list of Plutarchan allusions
to the HA in which Aristotle is not explicitly named, but it does not include the
present passage.

in this connection what applies to one of them applies to both, he
could have extracted a perfectly good explanation from the passages
of the Ethics to which his remarks apparently refer. There, Aristotle
offers the view that these failings arise only in connection with bodily
pleasures, whereas those that involve the soul are not subject to them
(though excessive indulgence in pleasures of the soul can properly be
criticized for other reasons, EN 1118a3–9); and it would have been
easy for Callistratus, armed with his thesis that the pleasures of sight
and hearing impinge on the soul, to point out that Aristotle’s restric-
tion of ἀκρασία and/or ἀκολασία to bodily pleasures will entail that
his own position is correct. In fact it will do so more straightforwardly
than it can on the basis of Aristotle’s way of distinguishing auditory
and visual pleasures from the others. It is therefore odd that he does
not make this rather obvious move.
But perhaps, in a sense, he does. Perhaps he assumes that the
relevant passages of the Ethics are so familiar that his well-educated
audience will have seen the point immediately, without any tedious
exposition of it from him. After all, in the first part of our passage,
Plutarch does not tell us that he is implicitly referring to Plato, and
still less does he expound the contents of the relevant bits of the
dialogues; he expects his readers to pick up the allusions for them-
selves. Callistratus could be pursuing the same strategy here. But there
is a difference, and it makes it rather more puzzling that he does not
make the crucial doctrine explicit. If, dull-witted as we are, we fail to
notice the Platonic allusions in the account of the musician’s behav-
iour, we can still read it as a perfectly coherent and intelligible
narrative. By contrast, if Callistratus’ remarks are not seen to be
underpinned by the Aristotelian thesis associating ἀκολαστία and/or
ἀκρασία with bodily pleasures and no others, they will apparently add
up to nothing at all. In that case it seems rather imprudent of him to
have left his audience to fill in the gap for themselves. Nevertheless he
does so, and in this case the implicit allusion makes an essential
contribution to our reading of the text.
Callistratus’ foray into Peripatetic territory ends at 705a8, and at
the beginning of the next sentence we find the connective ὅθεν,
‘hence’. What follows, then, is a collection of facts that can be
accounted for on the grounds that indulgence in auditory and visual
pleasures is immune to the charge of ἀκρασία. Specifically, it is this
that explains why people do not find it necessary to conceal their
susceptibility to these pleasures under the cloud of darkness, and why,
248 Andrew Barker

on the contrary, they build stadia and theatres in which they can be
enjoyed to the full and in public: ὅθεν οὐδεμία τῶν τοιούτων ἡδονῶν
ἀπόκρυφός ἐστιν οὐδὲ σκότους δεομένη καὶ τῶν τοίχων ‘περιθεόντων’, ὡς
οἱ Κυρηναϊκοὶ λέγουσιν, ἀλλὰ καὶ στάδια ταύταις καὶ θέατρα ποιεῖται
(705a9–b2). To see and hear things in company with many others,
Callistratus continues, is more pleasurable and honourable, σεμνότερον,
than doing so alone. Hence he can call on a multitude of witnesses
(μάρτυρας ἡμῶν ὅτι πλείστους λαμβανόντων) to the fact that indul-
gence in these pleasures is not a symptom of ἀκρασία or of patho-
logical susceptibility to pleasure, but is proper to a liberal and civilized
way of life (705b2–6).
Here we seem to have returned to the Academy. Plato’s references
to people who indulge pleasures of which they are ashamed, but do so
only in private, as at Laws 655d–656b, might be exploited to give
respectable authority to the use that Callistratus makes of these
points. But he gives the game away with his references to public
theatres and to the evidence of a great many witnesses, not to mention
his nod to the hedonistic Cyrenaics. The theatres point us to the
Athenian’s denunciations of popular ‘theatocracy’ in the Laws (700e–
701a), and the appeal to all these witnesses plainly reminds us of
Socrates’ comparisons of his method of argument with that of Polus
in the Gorgias (471e–472c, 473e–474b, 475e–476a); Polus thinks that
the more witnesses he can call on the better, whereas Socrates values
the testimony of one person only, the person with whom he is
currently in debate. But in any case we hardly need to hunt for
specific passages in the dialogues. The attitude that a Platonist of
any variety will take to the opinions of οἱ πόλλοι is entirely familiar,
and at least from a Platonist perspective, the observations that Calli-
stratus uses to illustrate the implications of his view and to provide
evidence in its support in fact undermine it fatally.
We might think of the continuous undercurrent of allusions in this
passage as a surreptitious running commentary on the written text,
a murmur of voices (especially Plato’s) in the background, passing
judgement on what Plutarch explicitly reports and guiding the
reader’s interpretation of it. Its role in the description of the musi-
cian’s performance is not strictly essential; it serves mainly to empha-
size and justify the unfavourable view of it which the text itself already
conveys. But Plutarch could hardly have given Callistratus a speech
which overtly subverted itself, and here the implicit allusions have a
much more important function, exposing the emptiness of his

arguments and destabilizing the thesis he champions, and so making
us ready to accept the contrary views which Lamprias expounds in
the sequel. And more generally, of course, if we acknowledge and
identify the ghostly presences that haunt the passage, we shall enrich
our appreciation of its meanings, and of the literary sophistication
with which Plutarch has composed it.21
INTERMEDIALITY
At the beginning of this chapter I suggested that the symposiasts’

reception of the speeches that follow the musician’s performance
should be thought of as mediated by their experience of the music
he performed and of their own responses to it. I also suggested that
something similar might be said of the readers of Plutarch’s dialogue,
or at any rate readers in Plutarch’s own era. Even though they were
not present to hear this particular performance, they are likely to
have had similar experiences of their own, memories which would
have been awakened by the highly coloured account presented here.
Music they had heard in the past, echoing in the ear of memory,
would have given substance to the descriptions in the text, and would
have been the principal reference-point for their interpretation of the
speeches. In this brief final section22 I would like to explore this
second suggestion a little further. In the closing paragraphs I shall
say a little—but only a little, since I can only skim the surface of the
concept here—about how it relates to the concept of intermediality
developed by theorists of various persuasions in recent decades.23
Consider, first, Plutarch’s account of the performance itself. It
conjures up vivid images, and eloquently expresses the attitude of
the writer, but at the same time it is curiously unspecific. The pieces of
music are not named or assigned to their composers; we are not even
told to which genres they belonged, only that the content of the first
part of the programme was admirable and that of the second part
21
For some further reflections on this passage see Barker (2016).
22
My thanks to Tom Phillips for encouraging me to engage with this issue, and for
his helpful comments on my attempts to do so.
23
For a valuable overview of intermediality, and a painstaking analysis of the
forms it can take, see Wolf (2002) which also contains an ample bibliography of earlier
studies.
250 Andrew Barker

disgraceful, provoking its audience to thoroughly unseemly behav-
iour. Uncertainties about the identity of the music are compounded
by the ambiguity of the word χορός, which we noticed earlier. Did
the music include contributions by a group of singers or dancers, or
indeed dancing singers? That seems to be the natural implication of the
allusion, but it is destabilized by the absence of any reference to choral
activity in the sequel. So perhaps the χορός is simply the musician’s fan-
club, like the χορός that attended Protagoras; but there’s nothing in the
text to confirm or confute that hypothesis either.
There is another uncertainty of an even more troublesome sort. At
the beginning of the passage the musician is described as an αὐλῳδός,
that is, a solo singer who performs to the accompaniment of an
aulos-player. But this identification fails to mesh convincingly with
Plutarch’s account of his performance. No accompanist is mentioned,
and the musician is not said to have conferred with anyone else at the
moment when he decides to abandon musical respectability. He
seems to be performing and taking decisions entirely on his own.
Nor is his activity described as singing. His deplorable music includes
κρούματα, a word that most often refers to the sounds of an instru-
ment but might be intended in the sense ‘rhythms’, and μέλη, which
means ‘melodies’ or ‘tunes’ regardless of whether they are vocal or
instrumental. But the third component regularly listed among the
ingredients of music, that is, the words, is obtrusively missing, despite
the fact that according to Plato the words of a song should take the
lead in determining the character of the music; the other elements in
the piece take their cue from them, and when music lacks words its
ethos is extremely hard to identify.24 Plutarch seems—whether delib-
erately or not—to be undermining his initial statement that the
musician is a singer. He has also planted in the text two hints which
positively encourage the reader to think of the performance as purely
instrumental; one is the reference to κρούματα, and the other is his
use of the verb καταυλεῖν to evoke the musician’s ‘spell-binding’ effect
on his audience. Neither of these settles the issue conclusively, since
24
Plato Rep. 398c–d; Laws 669e. In the same passage of the Republic, however,
Glaucon assigns a distinct ethos to each of the ἁρμονίαι in its own right, without
reference to any words that may be set to melodies formed within its framework
(398e–399c), and Socrates implies that the same could be done with rhythms (399e–
400c). Our passage of Plutarch implies that a complex made up of melody and rhythm
alone nevertheless has a determinate ethos, and that the absence of words does not
prevent us from detecting it.

both, like the allusion to a χορός, are infected with ambiguity.25 But
the idea that the musician is not an αὐλῳδός after all, but an instru-
mentalist and specifically an aulete, is further encouraged by the fact
that the most prestigious and famous musical event in the Pythian
contest was the competition for solo aulos-players, and if indications
to the contrary were lacking, we would naturally assume that it was
from this competition that the musician had been disqualified. In an
earlier version of this chapter I argued on grounds of these sorts that
Plutarch must have meant that the musician was an aulete, and that
the word αὐλῳδόν must therefore be an error, and should be emended
to αὐλητήν. As one of the anonymous readers of this volume has
pointed out to me, however, it isn’t easy to see how the corruption
could have happened, and I am now inclined to think that the
manuscript text should stand. In that case it seems impossible to
decide what sort of musician this is; we are pulled in one direction by
his explicit designation as an αὐλῳδός, and in the other by the cluster
of indications we have been reviewing.
To sum up these points, then, the ambiguities and apparent incon-
sistencies in Plutarch’s account leave the specifically musical aspects
of the performance indeterminate, and this indeterminacy opens up a
space to be filled out of the readers’ own resources. One might put it
more strongly: the text’s resistance to being pinned down to any
definite representation of the musician (or musicians), or of the
music that was performed, releases readers from the obligation to
imagine, at second hand, something they have not heard; it positively
invites them to bring their own experiences and memories to bear on
their interpretation of Plutarch’s brief narrative, and, more import-
antly, to keep them in mind as they read the subsequent speeches. No
matter what types of remembered music run through their heads
when they envisage the good and bad pieces discussed in the text, they
are at liberty to draw freely on them when assessing the merits of the
speakers’ arguments.
Readers’ memories of past experiences of music bring with them
memories of their own responses to these experiences. When they
consider Callistratus’ speech for the defence and the case for the
25
The sense of the verb καταυλεῖν is never far from that of enchanting or affecting
a person by means of aulos-playing, though it can also be used metaphorically to refer
to enchantments in which the aulos does not literally play a part, as e.g. at Plato
Rep. 411a.
252 Andrew Barker

prosecution subsequently presented by Lysias, these recollections
may prompt them, like the symposiasts themselves, either to feel
embarrassed at the shameful way they have reacted, or alternatively
to pride themselves on reacting in ways proper to a respectable
human being; and these stirrings of their consciences or their self-
esteem will modify their responses to the speakers’ arguments. Thus
Callistratus’ insistence that even non-rational creatures take pleasure
in music suggests that in this respect human beings are no different
from animals, and that their enjoyment of music is unconnected with
their distinctive rationality. Reflection on their own experiences
might lead readers to reject this implication, or to recognize its
cogency and—perhaps—to find it troubling. Similarly, Callistratus’
allusion to the behaviour of owls and the hunters who set out to catch
them invites the thought that music can lure us into a trap, with
disastrous consequences; and the fact that horse-breeders use music
to induce mares to submit to mating might be construed as a hint that
by by-passing our rational faculty it can entice us into acting in ways
contrary to our better judgement. Readers recalling their own musical
experiences might therefore be inclined to wonder whether they too
have sometimes fallen prey to such musical ‘entrapment’ and been
tempted into injudicious behaviour. In a rather different vein, Calli-
stratus’ appeal to the testimony of the multitudinous witnesses assem-
bled in the theatres might chime discordantly with their own
experiences. They might for instance recall a performance which
they had thought admirable, but to which the crowd in the theatre
had perversely responded with cat-calls and hails of rotten vegetables;
and anyone who now re-lived the disgust they had felt at the time for
these demonstrations of bad taste would be strongly disinclined to
follow Callistratus in putting his faith in the opinions of ‘the many’.
The preceding paragraph is littered with words such as ‘may’ and
‘might’, and this is inevitable. We can be sure that anyone present on
an occasion of the sort that Plutarch describes would have listened to
the speeches against the back-drop of the performance they had
heard. It’s also indisputable, I think, that the vagueness and ambigu-
ities of Plutarch’s description of the performance provide an open
opportunity for readers to fill the gaps in the way I have suggested.
Again, it seems certain that audiences for comedies in classical
Athens drew on their memories of other performances while they
listened; otherwise the comic dramatists’ musical spoofs and satirical
characterizations of musicians would have fallen on stony ground.

But if we ask whether remembered music was ringing in the ears of
contemporary readers of Plutarch’s dialogue, and whether they actu-
ally brought it to bear on their assessment of what they read, these are
questions we simply cannot answer. My suggestion that they ‘might’
have done so is a suggestion and no more.
Is it possible, finally, to relate these observations and suggestions to
the concept of intermediality, and if so, how? If we consider the
artifact with which Plutarch has presented us (that is, the written
text of the dialogue) simply by itself, it is clear that intermediality
plays no part at all. The text is not delivered through the medium of
music; nor does it modify its allusions to music by giving them—for
example—a rhythmic or metrical form reminiscent of those familiar
from musical compositions.26 What we find is something that is not
strictly a case of intermediality at all, but (in Wolf ’s terms) of ‘inter-
medial reference’; it refers to music, but without being ‘iconically
affected’ by its reference, or mediated through the filter of the music
it mentions. Wolf distinguishes two principal types of intermedial
reference. ‘Such reference . . . can point to another medium in general—
in which case it could be seen as a parallel to what in intertextuality
theory is called “system reference”; alternatively, intermedial reference
can also point to an individual work transmitted in another medium—in
which case the term “individual reference”, which is also derived from
intertextuality theory, would be applicable.’27
The distinction between system reference and individual reference,
however, is less clear-cut than it might appear. Imagine a scene in a
novel in which someone walks along a corridor and hears music
coming through an open door. If (a) we are told only that the
character hears music, we are evidently in the presence of system
reference. But the author might have made the reference more spe-
cific: (b) the music is a piano concerto, or more specifically still,
(c) one of Mozart’s piano concertos. Still further along the continuum
between system reference and individual reference, the author might
have chosen to tell us (d) that it is Mozart’s C Major piano concerto,
or again (e) that it is the slow movement of that concerto; or it might
be identified even more closely (f) as the plangent descending
sequence that recurs thematically in the concerto’s slow movement.
26
This would be what Werner Wolf calls ‘intermedial imitation’: Wolf (2002)
24–5.
27
Wolf (2002) 23.
254 Andrew Barker

If the reference the novel provides takes the form of (d), (e), or (f),
some of the novel’s readers will be familiar with this concerto and
some will not. The responses of the latter to the text are bound to be
unaffected by past experiences of the music itself, but those of
the former might very well be so affected, and the more specific the
reference becomes, the more likely it becomes that memories of
the music will intervene in their reception of what they read. If the
novelist’s allusion takes the form of (f), it is in fact hard to believe that
readers who know the music well will be able to resist recreating the
sequence in their mental ear, thereby enriching the atmosphere of the
scene that the novel depicts. When they do so, a form of intermedi-
ality evidently comes into play; and since its intervention depends on
a contribution which the reader may or may not provide, and is not
inherent in anything actually contained in the novel, we might dis-
tinguish it from other forms by calling it ‘potential intermediality’.
Although the dialogue’s references to music are not strictly parallel
to any of (a), (b), and (c), they are closer to the ‘system reference’ end
of the spectrum than the ones we have just been considering. If the
notion of potential intermediality is applicable in their case too
(which is roughly what I have suggested), it must be applied in a
rather different way. Plutarch’s account of the episode in which both
good and bad music are performed offers some clues about what the
music was like, but the vagueness and ambiguities we noted above
leave most of the details open. Some readers will have had experiences
which they could call upon to fill in the gaps, and some, perhaps, will
not; and of those who have had such experiences, some will reactivate
them in their memories as they read and some will not. Those who do
so will, in effect, be converting Plutarch’s ‘system references’ into
‘individual references’, removing the generality of Plutarch’s descrip-
tions by identifying the referents with particular instances that fall
within their scope. In so doing, they, like their counterparts among
the readers of our hypothetical novel, will bring an intermedial
dimension to their engagement with and appreciation of the text. It
therefore seems as legitimate to locate potential intermediality in the
Plutarchan dialogue as it is to do so in the scene in the novel. But the
two cases differ in one significant respect. The text of the novel
determines the identity of the music through which its potential
intermediality can be actualized, and no other music will serve the
purpose. By contrast, Plutarch’s text indicates no more than the
music’s general characteristics, leaving its identification with any

individual compositions or performances to be determined by the
reader. The ‘potential’ which readers may or may not actualize is
therefore of a looser sort, and it resides in them as much as it does in
anything Plutarch has written. Nevertheless, insofar as the text admits
the possibility of intermedial intervention in a reader’s reception of
it—and indeed, in my view, positively invites such intervention—the
concept of potential intermediality is applicable here too; and by
recognizing this fact we can enlarge our understanding of the
range of responses to the dialogue that were available to Plutarch’s
contemporaries.
Bibliography
Abbenes, J., Slings, S., and Sluiter, I. (eds) (1995) Greek Literary Theory after
Aristotle, Amsterdam.
Adler, A. (ed.) (1928–38) Suidae Lexicon, i–v, Leipzig.
Agócs, P., Carey, C., and Rawles, R. (eds) (2012) Reading the Victory Ode,
Cambridge.
Alcock, S., Cherry, J., and Elsner, J. (2001) Pausanias: Travel and Memory in
Roman Greece, Oxford.
Allan, W. (2008) Euripides, Helen, Cambridge.
Anderson, W. D. (1966) Ethos and Education in Greek Music, Cambridge, MA.
Anderson, W. D. (1980) ‘Mesomedes’ in Stanley Sadie (ed.), The New Grove
Dictionary of Music and Musicians, London.
Arnott, W. G. (2007) Birds in the Ancient World from A to Z, London.
Aron, K. (1920) ‘Beiträge zu den Persern des Timotheos’, PhD diss.,
Erlangen.
Asmis, E. (1992) ‘Crates on Poetic Criticism’, Phoenix 46, 128–69.
Athanassaki, L. (2009) ‘Narratology, Deixis, and the Performance of Choral
Lyric: On Pindar’s First Pythian Ode’, in J. Grethlein and A. Rengakos
(eds), 241–73.
Athanassaki, L. and Bowie, E. (eds) (2011) Archaic and Classical Choral Song,
Berlin.
Aubenque, P. and Tordesillas, A. (eds) (1993) Aristote Politique, Paris.
Austin, M. M. (2005) «Κοινωνία καὶ οἰκονομία», in Πανεπιστήμιο τοῦ
Καῖμπριτζ, Ἱστορία τῆς ἀρχαίας Ἑλλάδας 6 (tr. Εὐάγγελος Χεμίκογλου
et al.), 208–47, Athens.
Bachvarova, M. R. (2016) From Homer to Hittite, Cambridge.
Barker, A. (1984) Greek Musical Writings, vol. 1, Cambridge.
Barker, A. (1989) Greek Musical Writings, vol. 2, Cambridge.
Barker, A. (2002) ‘La musica di Mesomede’, in F. Perusino and E. Rocconi
(eds), Euterpe, 133–48, Pisa.
Barker, A. (2007) The Science of Harmonics in Classical Greece, Cambridge.
Barker, A. (2014) Ancient Greek Writers on Their Musical Past, Rome.
Barker, A. (2016) ‘Plutarch, Quaestiones convivales 704c4–705b6: The Host
and the Musician’, in L. Bravi, L. Lomiento, A. Meriani, and G. Pace (eds),
15–28.
Baughan, E. P. (2013) Couched in Death: Klinai and Identity in Anatolia and
Beyond, Madison, WI.
Bayer, B. (2014) ‘The Mesopotamian Theory of Music and the Ugarit Notation:
A Reexamination’, in Y. Maurey, E. Seroussi, and J. Westenholz (eds), 15–91.
258 Bibliography
Beck, M. (ed.) (2014) A Companion to Plutarch, Chichester.
Beecroft, A. J. (2008) ‘Nine Fragments in Search of an Author: Poetic Lines
Attributed to Terpander’, CJ 103, 225–42.
Bélis, A. (2003a) ‘Un lyrikos de l’époque des Antonins: Mésomède de Crète’,
in J. Jouanna and J. Leclant (eds), 223–35.
Bélis, A. (2003b) ‘Le Péan de Berlin: une reflecture’, Revue des études grecques
116.2, 537–58.
Benedetti, F. and Grandolini, S. (eds) (2003) Studi di filologia e tradizione
greca in memoria di Aristide Colonna, Naples.
Berman, D. W. (2007) Myth and Culture in Aeschylus’ Seven against Thebes,
Rome.
Bernhart, W. (ed.) (2012) Word and Music Studies: Essays on Performativity
and on Surveying the Field, Amsterdam.
Bernhart, W., Scher, S., and Wolf, W. (eds) (1997) Word and Music Studies:
Defining the Field, Amsterdam and Atlanta.
Bispham, E. and Harrison, T. (eds) (2006) Edinburgh Companion to Greece
and Rome, Edinburgh.
Blank, D. L. (1982) Ancient Philosophy and Grammar: The Syntax of Apol-
lonius Dyscolus, Chico, CA.
Boardman, John (1956) ‘Some Attic Fragments: Plot, Plaque, and Dithyramb’,
JHS 76, 18–24.
Böhme, G. (1993) ‘Atmosphere as the Fundamental Concept of a New
Aesthetics’, Thesis Eleven 36, 113–26.
Braund, D. and Wilkins, J. (eds) (2000) Athenaeus and His World, Exeter.
Bravi, L., Lomiento, L., Meriani, A., and Pace, G. (eds) (2016) Tra Lyra e
Aulos, Pisa and Rome.
Buchan, M. (2008) ‘ “Too Difficult for a Single Man to Understand”: Medea’s
Out-Jutting Foot’, Helios 35, 3–28.
Budelmann, F. (ed.) (2009) The Cambridge Companion to Greek Lyric,
Cambridge.
Bull, M. and Back, L. (eds) (2003) The Auditory Culture Reader, Oxford.
Burkert, W. (1972) Lore and Science in Ancient Pythagoreanism, Cambridge,
MA.
Burkert, W. (1992) The Orientalizing Revolution, Cambridge, MA.
Burnett, A. P. (1984) Review of W. Mullen, Choreia: Pindar and Dance, CP
79, 154–60.
Butler, S. (2015) The Ancient Phonograph, Cambridge, MA and London.
Butler, S. and Nooter, S. (eds) (2017) Sound and the Ancient Senses, London.
Calame, C. (2001) Choruses of Young Girls in Ancient Greece, London.
Caplin, W. E. (1998) Classical Form: A Theory of Formal Functions, Oxford
and New York.
Capra, A. (2014) Plato’s Four Muses: The Phaedrus and the Poetics of
Philosophy, Washington, DC.
Bibliography 259
Castaldo, D., Giannachi, F., and Manieri, A. (eds) (2012) Poesia, musica e
agoni nella Grecia antica, Galatina.
Cassio, A., Musti, D., and Rossi, L. (eds) (2000) Synaulía: cultura musicale in
grecia e contatti mediterranei, Naples.
Catenacci (2013) = Catenacci, C., Giannini, P., and Lomiento, L. (eds)
Pindaro Le Olimpiche, Milan.
Celentano, M. S. (ed.) (2010) Ricerche di metrica e musica greca per Roberto
Pretagostini, Alexandria.
Cheng, J. (2001) ‘Assyrian Music as Represented and Representations of
Assyrian Music’, PhD Diss., Harvard.
Chiasson, C. (1988) ‘Lecythia and the Justice of Zeus in Aeschylus’ Oresteia’,
Phoenix 42, 1–21.
Coker, W. (1972) Music and Meaning: A Theoretical Introduction to Musical
Aesthetics, New York.
Collard, C. and Morwood, J. (2017) Euripides: Iphigenia in Aulis.
Warminster.
Collins, B., Bachvarova, M., and Rutherford, I. (eds) (2008) Anatolian Inter-
faces: Hittites, Greeks and Their Neighbors in Ancient Anatolia, Oxford.
Collins, D. (1999) ‘Hesiod and the Voice of the Muses’, Arethusa 32.3,
241–62.
Comotti, G. (1972) ‘L’endecacorde di Ione di Chio’, QUCC 13, 54–61.
Cook, N. (2001) ‘Theorizing Musical Meaning’, Music Theory Spectrum 23.2,
170–95.
Cooke, D. (1959) The Language of Music, Oxford.
Cosgrove, C. H. and Meyer, M. C. (2006) ‘Melody and Word-Accent Rela-
tionships in Ancient Greek Musical Documents: The Pitch Height Rule’,
JHS 126, 66–81.
Costil, P. (1949) ‘L’Esthétique littéraire de Denys d’Halicarnasse’. Diss.,
Sorbonne.
Cousland, J. and Hume, J. (eds) (2009) The Play of Texts and Fragments,
Leiden.
Crane, G. (1990) ‘A Change of Fashion and Euripides’ Medea 190–203’,
Mnem. 43, 435–8.
Creese, D. (2010) The Monochord in Ancient Greek Harmonic Science,
Cambridge.
Cribiore, R. (2001) Gymnastics of the Mind: Greek Education in Hellenistic
and Roman Egypt, Princeton, NJ.
Crocker, R. L. (1978) ‘Remarks on the Tuning Text UET VII 74 (U. 7/80)’,
Or. 47, 99–104.
Crocker, R. L. (1997) ‘Mesopotamian Tonal Systems’, Iraq 59, 189–202.
Cropp, M. (1988) Euripides Electra, Warminster.
Cropp, M. and Fick, G. (1985) Resolutions and Chronology in Euripides: the
Fragmentary Tragedies, London.
260 Bibliography
Cropp, M., Lee, K., and Sansone, D. (eds) (2000) Euripides and Tragic
Theatre in the Late Fifth Century, Champaign, IL.
Crusius, O. (1888) ‘Stesichorus and die Epodische Composition in der
Griechischen Lyrik’, Commentationes Ribbeckianae, Leipzig, 1–22.
Csapo, E. (1999–2000) ‘Later Euripidean Music’, M. Cropp, K. Lee, and
D. Sansone (eds), 399–426.
Csapo, E. (2003) ‘The Dolphins of Dionysus’ in E. Csapo and M. Miller (eds),
69–98.
Csapo, E. (2004) ‘The Politics of the New Music’, in P. Murray and P. Wilson
(eds), 207–48.
Csapo, E. (2008) ‘Star Choruses: Eleusis, Orphism, and New Musical
Imagery and Dance’ in M. Revermann and P. Wilson (eds), 262–90.
Csapo, E. (2009) ‘New Music’s Gallery of Images: The “Dithyrambic” First
Stasimon of Euripides’ Electra’ in J. Cousland and J. Hume (eds), 95–109.
Csapo, E. (2011) ‘Economics, Poetics, Politics, Metaphysics, and Ethics of the
“New Music”’, in D. Yatromanolakis (ed.) (2011), 65–132.
Csapo, E. and Miller, M. (eds) (2003) Poetry, Theory, Praxis, Oxford.
Csapo, E. and Wilson, P. (2009) ‘Timotheus the New Musician’ in
F. Budelmann (ed.), 277–94.
Culler, J. (2015) Theory of the Lyric, Cambridge, MA and London.
Currie, B. (2012) ‘Pindar and Bacchylides’ in I. de Jong (ed.), 285–303.
D’Angour, A. J. (1997) ‘How the Dithyramb Got Its Shape’, CQ 47, 331–51.
D’Angour, A. J. (1999) ‘Archinus, Eucleides, and the reform of the Athenian
alphabet’, BICS 43, 109–30.
D’Angour, A. J. (2005) ‘Intimations of the Classical in Early Greek mousikē’
in J. Porter (ed.), 89–105.
D’Angour, A. J. (2006a) ‘The New Music: So What’s New?’ in S. Goldhill and
R. Osborne (eds), 264–83.
D’Angour, A. J. (2006b) ‘Metre’, in E. Bispham and T. Harrison (eds), 489–94.
D’Angour, A. J. (2007) ‘The Sound of Mousikē: Reflections on Aural Change
in Ancient Greece’, in R. Osborne (ed.), 288–300.
D’Angour, A. J. (2011) The Greeks and the New: Novelty in Ancient Greek
Imagination and Experience, Cambridge.
D’Angour, A. J. (2013) ‘Music and Movement in the Dithyramb’, in
B. Kowalzig and P. Wilson (eds), 198–210.
D’Angour, A. J. (2015) ‘Sense and Sensation in Music’, in P. Destrée and
P. Murray (eds), 188–203.
D’Angour, A. J. (2017) ‘Euripides and the Sound of Music’, in L. McClure
(ed.), 428–43.
Dale, A. M. (1968) The Lyric Metres of Greek Drama (2nd edn), Cambridge.
Dale, A. M. (1969) Collected Papers, Cambridge.
Danek, G. and Hagel, S. (1995) ‘Homer-Singen’, Wiener Humanistische
Blätter 37, 5–20.
Bibliography 261
Davies, M. and Finglass, P. (2014) Stesichorus: The Poems, Cambridge.
Davies, S. (1994) Musical Meaning and Expression, Ithaca, NY.
Demont, P. (1993) ‘Le loisir dans la Politique d’Aristote’, in P. Aubenque and
A. Tordesillas (eds), 209–30.
Demos, M. (1999) Lyric Quotation in Plato, Lanham, MD.
Depew, D. J. (1991) ‘Politics, Music, and Contemplation in Aristotle’s Ideal
State’, in D. Keyt and F. Miller, Jr (eds), 346–80.
Destrée, P. (2017) ‘Aristotle and Musicologists on Three Functions of
Music: A Note on Pol. 8, 1341b40–1’, GRMS 5, 35–42.
Destrée, P. and Murray, P. (eds) (2015) A Companion to Ancient Aesthetics,
Chichester.
Deubner, L. (1929) ‘Die viersaitige Leier’, Athenische Mitteilungen 54, 194–200.
Deubner, L. (1930) ‘Terpander und die siebensaitige Leier’, PhW 50, 1566–7.
Dillon, E. (2012) The Sense of Sound: Musical Meaning in France, 1260–1330,
Oxford.
Dobree, P. P. (1833) Petri Pauli Dobree tomus posterior, Cambridge.
Dolar, M. (2006) A Voice and Nothing More, Cambridge, MA.
Dougherty, C. and Kurke, L. (eds) (2003) The Cultures Within Ancient Greek
Culture, Cambridge.
Dreyfus, L. (2013) ‘Allusive Representations: Homoerotics in Wagner’s
Tristan’, in J. Walden (ed.), 167–81.
Duchesne-Guillemin, M. (1969) ‘La Theorie babylonienne des métaboles
musicales’, Rev. Mus. 55, 3–11.
Easterling, P. E. (ed.) (1997) The Cambridge Companion to Greek Tragedy,
Cambridge
Easterling, P. and Knox, B. (1999) Ἱστορία τῆς ἀρχαίας ἑλληνικῆς λογοτεχνίας
(tr. Ν. Κονομής, Χρ. Γρίμπα, and Μ. Κονομή) (3rd edn), Athens.
Ebeling, E. (1919) Keilschrifttexte aus Assur religiösen Inhalts, Leipzig.
Ebeling, E. (1922) ‘Einen Hymnen-Katalog aus Assur’, Berliner Beiträge zur
Keilschriftforschung 1.3, 1–32.
Ebeling, E. and Meissner, B. (eds) (1997) Reallexikon der Assyriologie und
vorderasiatischen Archäologie 8, Berlin.
Edmonds, J. M. (1922) ‘Sappho’s Book as Depicted on an Attic Vase’, CQ 16,
1–14.
Edwards, M. W. (2002) Sound, Sense, and Rhythm: Listening to Greek and
Latin Poetry, Princeton, NJ.
England, E. B. (1921) The Laws of Plato: Books One to Six, Manchester.
Ercoles, M. (2010) ‘Note a Tim. PMG 791, 221–228’, Eikasmos 21, 111–32.
Faraone, C. (2006) The Stanzaic Architecture of Early Greek Elegy, Oxford.
Färber, H. (1936) Die Lyrik in der Kunsttheorie der Antike, Munich.
Fearn, D. (ed.) (2011) Aegina: Contexts for Choral Lyric Poetry, Oxford.
Ferrari, G. (1990) Listening to the Cicadas: A Study of Plato’s Phaedrus, Cambridge.
Ferrari, G. (2008) Alcman and the Cosmos of Sparta, Chicago.
262 Bibliography
Ferreira, J. R., Leão, D., Tröster, M., and Dias, P. B. (eds) (2009) Symposion
and Philanthropia in Plutarch, Humanitas Suppl. 6, Coimbra.
Finglass, P., Collard, C., and Richardson, N. (eds) (2007) Hesperos: Studies in
Ancient Greek Poetry Presented to M. L. West on His Seventieth Birthday,
Oxford.
Finkel, I. and Civil, M. (1982) The Series SIG ALAN = Nabnitu, Rome.
Finkelberg, M. (ed.) (2011) The Homer Encyclopedia, Oxford.
Fitzgerald, W. (1987) Agonistic Poetry, Berkeley, Los Angeles, and London.
Flintoff, E. (1986) ‘The date of the Prometheus Bound’, Mnem. 39, 82–91.
Ford, A. (1992) Homer: The Poetry of the Past, Ithaca, NY.
Ford, A. (2004) ‘Catharsis: The Power of Music in Aristotle’s Politics’ in
P. Murray and P. Wilson (eds), 309–36.
Ford, A. (2013) ‘The Poetics of Dithyramb’, in B. Kowalzig and P. Wilson
(eds), 313–31.
Fox-Strangways, A. H. (1914) The Music of Hindostan, Oxford.
Franklin, J. C. (2002a) ‘Harmony in Greek and Indo-Iranian Cosmology’,
JIES 30.1–2, 1–25.
Franklin, J. C. (2002b) ‘Terpander: The Invention of Music in the Oriental-
izing Period’, PhD Diss., London.
Franklin, J. C. (2002c) ‘Musical Syncretism and the Greek Orientalizing
Period’, in E. Hickmann, R. Eichmann, Kilmer and A. (ed.), Archäologie
früher Klangerzeugung und Tonordnungen, 441–51, Rahden.
Franklin, J. C. (2002d) ‘Diatonic Music in Greece: A Reassessment of its
Antiquity’, Mnem. 56.1, 669–702.
Franklin, J. C. (2003) ‘The Language of Musical Technique in Greek Epic
Diction’, Gaia 7, 295–307.
Franklin, J. C. (2004) ‘Structural Sympathies in Ancient Greek and South-
Slavic Heroic Song’, in E. Hickmann and R. Eichmann (eds), 241–51.
Franklin, J. C. (2005) ‘Hearing Greek Microtones’, in S. Hagel and
C. Harrauer (eds), Ancient Greek Music in Performance, 9–50, Vienna.
Franklin, J. C. (2006) ‘Lyre Gods of the Bronze Age Musical Koine’, JANER
6.2, 39–70.
Franklin, J. C. (2007) ‘The Global Economy of Music in the Ancient Near
East’, in J. G. Westenholz (ed.), 27–37.
Franklin, J. C. (2008) ‘A Feast of Music’: the Greco-Lydian Musical Move-
ment on the Assyrian Periphery’, in B. Collins, M. Bachvarova, and
I. Rutherford (eds), 193–204.
Franklin, J. C. (2010) ‘Remembering Music in Early Greece’, in S. Mirelman
(ed.), 9–50.
Franklin, J. C. (2011) ‘Music’, in M. Finkelberg (ed.), The Homer Encyclopedia,
Oxford.
Franklin, J. C. (2012) ‘The Lesbian Singers: Towards a Reconstruction of Hellanicus’
“Karneian Victors”’, in D. Castaldo, F. Giannachi, and A. Manieri (eds), 720–64.
Bibliography 263
Franklin, J. C. (2013) ‘Song-Benders of Circular Choruses’: Dithyramb and
the “Demise of Music” ’, in P. Wilson and B. Kowalzig (eds), 213–36.
Franklin, J. C. (2016) Kinyras: The Divine Lyre, Washington, DC.
Fulcher, J. F. (2001) Debussy and His World. Princeton, NJ.
Gagné, R. (2013) ‘Dancing Letters: The Alphabetic Tragedy of Kallias’, in
R. Gagné and M. G. Hopman (eds), 297–316.
Gagné, R. and Hopman, M. G. (eds) (2013) Choral Mediations in Greek
Tragedy, Cambridge.
Gantz, T. (1993) Early Greek Myth: A Guide to Literary and Artistic Sources,
Baltimore.
Gentili, B. (1988) Poetry and Its Public in Ancient Greece, Baltimore.
Gentili, B. and Perusino, F. (eds) (1995) Mousike: metrica ritmica e musica
greca: in memoria di Giovanni Comotti, Pisa and Rome.
Gentili, B. and Pretagostini, R. (eds) (1988) La musica in Grecia, Rome and
Bari.
Goldhill, S. (2013) ‘Choreography: the Lyric Voice of Sophoclean Tragedy’,
in R. Gagné and M. G. Hopman (eds), 100–29.
Goldhill, S. and Osborne, R. (eds) (1999) Performance Culture and Athenian
Democracy, Cambridge.
Goldhill, S. and Osborne, R. (eds) (2006) Rethinking Revolutions through
Ancient Greece, Cambridge.
Gostoli, A. (1990) Terpandro: Introduzione, testimonianze, testo critico,
traduzione e commento, Rome.
Greene, E. (ed.) (2005) Women Poets in Ancient Greece and Rome, Norman,
OK.
Gregory, J. (ed.) (2008) A Companion to Greek Tragedy, Cambridge.
Grethlein, J. and Rengakos, A. (eds) (2009) Narratology and Interpretation,
Berlin and New York.
Griffith, M. (1977) The Authenticity of Prometheus Bound, Cambridge.
Griffith, M. (2013) ‘Satyr-Play, Dithyramb, and the Geopolitics of Dionysian
Style in Fifth-Century Athens’, in B. Kowalzig and P. Wilson (eds),
257–81.
Groddeck, W. (1995) Reden über Rhetorik: Zu einer Stilistik des Lesens, Basel.
Gülgönen, S. (2011) ‘L’éducation musicale dans les lois’, Journal of Ancient
Philosophy 5, 1–16.
Gumbrecht, H. U. (2012) Atmosphere, Mood, Stimmung: On a Hidden
Potential of Literature (tr. E. Butler), Stanford.
Gurd, S. (2016) Dissonance: Auditory Aesthetics in Ancient Greece, New
York.
Gurney, O. R. (1968) ‘An Old Babylonian Treatise of the Tuning of the
Harp’, Iraq 30, 229–33.
Gurney, O. R. (1974) Ur Excavation Texts 7, London.
Gurney, O. R. (1994) ‘Babylonian Music Again’, Iraq 56, 101–6.
264 Bibliography
Gurney, O. R. and West, M. L. (1998) ‘Mesopotamian Tonal Systems:
A Reply’, Iraq 60, 223–7.
Güterbock, H. and Jacobsen, T. (eds) (1965) Studies in Honor of Benno
Landsberger, Chicago.
Andújar, R., Coward, T., and Hadjimichael, T. (eds) (forthcoming) Paths of
Song: Interactions between Greek Lyric and Tragedy, Berlin.
Hagel, S. (1994) ‘Zu den Konstituenten des griechischen Hexameters’, WS
107/8, 77–108.
Hagel, S. (2000) Modulation in altgriechischer Musik: Antike Melodien im
Licht antiker Musiktheorie, Frankfurt.
Hagel, S. (2005a) ‘Twenty-Four in Auloi: Aristotle, Met. 1093b, the Harmony
of the Spheres, and the Formation of the Perfect System’, in S. Hagel and
C. Harrauer (eds), 51–92.
Hagel, S. (2005b) ‘Is Nid Qabli Dorian?’, Bagh. Mitt. 36, 287–348.
Hagel, S. (2006) ‘The Context of Tunings: Thirds and Septimal Intervals in
Ancient Greek Music’, in E. Hickmann, A. Both, and R. Eichmann (eds),
281–304.
Hagel, S. (2009) Ancient Greek Music: A New Technical History, Cambridge.
Hagel, S. (2012) ‘The aulos syrinx’ in D. Castaldo, F. Giannachi, and
A. Manieri (eds), 489–518.
Hagel, S. (1994) ‘Zu den Konstituenten des griechischen Hexameters’, Wien.
Stud. 107/108, 77–108.
Hagel, S. and Harrauer, C. (eds) (2005) Ancient Greek Music in Performance,
Vienna.
Hall, E. (1997) ‘The Sociology of Athenian Tragedy’, in P. E. Easterling (ed.),
93–126.
Hall, E. (1999) ‘Actor’s Song in Tragedy’, in S. Goldhill and R. Osborne (eds),
96–122.
Hall, E. (2006) The Theatrical Cast of Athens, Oxford.
Halliwell, S. (2002) The Aesthetics of Mimesis, Princeton, NJ.
Hanslick, E. (1974) The Beautiful in Music: A Contribution to the Revisal of
Musical Aesthetics, New York.
Harrison, G. (ed.) (2005) Satyr Drama: Tragedy at Play, Swansea.
Heimpel, W. (2015) ‘Balang Gods’, in J. C. Franklin, Kinyras: The Divine
Lyre, 571–632, Washington, DC.
Held, M. (1961) ‘A Faithful Lover in an Old Babylonian Dialogue’, JCS 15,
1–26.
Henderson, C., Jr (ed.) (1964) Classical, Mediaeval, and Renaissance Studies
in Honor of B. L. Ullman, vol. 1, Rome.
Hense, O. (1876) ‘Die ABC-tragödie des Kallias und die Medea des Euripi-
des’, RhM 31, 582–601.
Hertz, D. M. (1987) The Tuning of the World: The Musico-Literary Poetics of
the Symbolist Movement, Carbondale, IL.
Bibliography 265
Hickmann, E., Both, A., and Eichmann, R. (eds) (2006) Musikarchäologie im
Kontext: Archäologische Befunde, historische Zusammenhänge, soziokul-
turelle Beziehungen, Rahden.
Hickmann,E.and Eichmann, R.(eds) (2004) Music Archeological Sources,Rahden.
Hoffmann, E. T. A. (2001) Die Serapions-Brüder, in W. Segebrecht and
H. Steinecke (eds), Sämtliche Werke 4, Frankfurt.
Hopman, M. G. (2008) ‘Revenge and Mythopoiesis in Euripides’ Medea’,
TAPA 138, 155–83.
Horden, P. (ed.) (2000) Music as Medicine, Aldershot.
Hordern, J. (2002) The Fragments of Timotheus of Miletus, Oxford.
Hornblower, S. and Morgan, C. (eds) (2007) Pindar’s Poetry, Patrons and
Festivals, Oxford.
Horrocks, G. (2010) Greek: A History of the Language and Its Speakers (2nd
edn), Cambridge.
Huffman, C. (ed.) (2012) Aristoxenus of Tarentum, 129–54, New Brunswick
and London.
Hunter, R. and Rutherford, I. (eds) (2009) Wandering Poets in Ancient Greek
Culture, Cambridge.
Hüschen, H. and Moser, D.-R. (eds) (1974) Convivium Musicorum: Fest-
schrift Wolfgang Boetticher, Berlin.
Ihde, D. (2003) ‘Auditory Imagination’, in M. Bull and L. Back (eds), 61–6.
Immerwahr, H. R. (1964) ‘Book Rolls on Attic Vases’, in C. Henderson, Jr
(ed.), 17–48.
Irigoin, J. (1988) ‘Le premier stasimon de la Médée d’Euripide (v.410–45):
Répétitions de mots et échos de sonorités’, in Mélanges de la Bibliothèque
de la Sorbonne offerts à André Tuilier, Paris, 11–18.
Itsumi, K. (2009) Pindaric Metre: The ‘Other Half ’, Oxford.
Jaeger, C. S. (ed.) (2010) Magnificence and the Sublime in Medieval Aesthetics,
New York.
Jan, C. von (1962 [1895]) Musici Scriptores Graeci, Hildesheim.
Jan, C. von (1894) ‘Die Harmonie der Sphären’, Philologus 52, 13–37.
Jankelevitch, M. (2003) Music and the Ineffable, Princeton, NJ and Oxford.
Janko, R. (2000) Philodemus, On Poems, Oxford.
Janssen, T. H. (1984) Timotheus Persae: A Commentary, Amsterdam.
Jennings, V. and Katsaros, A. (eds) (2007) The World of Ion of Chios, Leiden
and Boston.
de Jong, I. (2012) (ed.) Space in Ancient Greek Literature, Leiden.
Jones, E. M. (2012) ‘Allocating Musical Pleasure: Performance, Pleasure, and
Value in Aristotle’s Politics’ in I. Sluiter and R. Rosen (eds), 159–82.
Jouanna, J. and Leclant, J. (eds) (2003) Colloque La poésie grecque antique.
Actes. (Cahiers de la villa Kérylos, 14), Paris.
Jung-Kaiser, U. (1997) ‘Die wahren Bilder und Chiffren “tragischer Ironie”
in Mahlers “Erster” ’, in G. Weiss (ed.), Neue Mahleriana, 101–52, Bern.
266 Bibliography
Kafka, F. (1994) Drucke zu Lebzeiten, eds. H.-G. Koch, M. Müller, and
M. Pasley, Frankfurt.
Kaibel, G. (1923–5) Athenaei Naucratitae Dipnosophistarum libri xv, Leipzig.
Kakrides, I. T. (ed.) (1986a) Ἑλληνικὴ μυθολογία I: Εἰσαγωγὴ στὸ μῦθο,
Athens.
Kakrides, I. T. (1986b) Ἑλληνικὴ μυθολογία II: Οἱ θεοί, Athens.
Kattenbelt, C. (2008) ‘Intermediality in Theatre and Performance: Defin-
itions, Perceptions and Medial Relationships’, Cultural Studies Journal of
Universitat Jaume I 6, 19–29.
Keen, A. (2009) ‘ “Undoing the Wineskin’s Foot”: Athenian Slang?’, CQ 59,
626–31.
Keyt, D. and Miller, F., Jr (eds) (1991) A Companion to Aristotle’s Politics,
Oxford.
Kilmer, A. D. (1960) ‘Two New Lists of Key Numbers for Mathematical
Operations’, Or. 39: 273–308.
Kilmer, A. D. (1965) ‘The Strings of Musical Instruments: Their Names,
Numbers and Significance’, in H. Güterbock and T. Jacobsen (eds), 261–8.
Kilmer, A. D. (1971) ‘The Discovery of an Ancient Mesopotamian Theory of
Music’, PAPhS 115, 131–49.
Kilmer, A. D. (1997) ‘Musik A. I. in Mesopotamien’, in E. Ebeling and
B. Meissner (eds), 463–82.
Kirk, G. S. (1985) The Iliad: A Commentary, vol. 1, Cambridge.
Kivy, P. (2007) Music, Language, and Cognition, Oxford.
Klotz, F. (2014) ‘The sympotic works’, in M. Beck (ed.), 207–22.
Klotz, F. and Oikonomopoulou, K. (eds) (2011) The Philosophers’ Banquet,
Oxford.
Koller, H. (1956) ‘Die Parodie’, Glotta 35, 17–32.
Koller, H. (1962) ‘Dithyrambos und Tragödie’, Glotta 40, 183–95.
Koopman, C. and Davies, S. (2001) ‘Musical Meaning in a Broader Perspec-
tive’, Journal of Aesthetics and Art Criticism 59.3, 261–73.
Kovacs, D. (1994) Euripides: Cyclops, Alcestis, Medea (Loeb), Cambridge,
MA.
Kovacs, D. (2002) Euripides Bacchae, Iphigenia at Aulis, Rhesus (Loeb),
Cambridge, MA.
Kowalzig, B. (2007) Singing for the Gods, Oxford.
Kowalzig, B. (2013a) ‘Dancing Dolphins on the Wine-Dark Sea: Dithyramb
and Social Change in the Archaic Mediterranean’, in B. Kowalzig and
P. Wilson (eds), 31–58.
Kowalzig, B. (2013b) ‘Broken Rhythms in Plato’s Laws’, in A.-E. Peponi (ed.),
171–211.
Kowalzig, B. and Wilson, P. (eds) (2013) Dithyramb in Context, Oxford.
Kramer, L. (1990) Music as Cultural Practice, Berkeley.
Kraut, R. (1997) Aristotle: Politics Books VII and VIII, Oxford.
Bibliography 267
Krispijn, T. J. H. (1990) ‘Beitrage zur altorientalischen Musikforschung 1.
Šulgi und die Musik’, Akkadica 70, 1–27.
Kritzas, C. B. (1980) ‘Muses delphiques à Argos’, BCH Supplement 6,
195–209.
Kroll, W. (1907) ‘Randbemerkungen’, RhM 62, 86–101.
Kroll, W. (ed.) (1913) M. Tullii Ciceronis Orator, Berlin.
Kümmel, H. M. (1970) ‘Zur Stimmung der babylonischen Harfe’, Or. 39,
252–63.
Lamberton, R. and Keaney, J. (eds) (1992) Homer’s Ancient Readers,
Princeton, NJ.
Landels, J. G. (1999) Music in Ancient Greece and Rome, London and New
York.
Langer, S. K. (1976) Philosophy in a New Key (3rd edn), Cambridge, MA.
Lasserre, F. (1988) ‘Musica babilonese e musica greca’, in B. Gentili and
R. Pretagostini (eds), 72–83.
Latacz, J. (1968) ‘Apteros muthos, apteros phatis: ungeflügelte Worte?’, Glotta
46, 27–47.
Lawler, L. (1954) ‘Phora, Schêma, Deixis in the Greek Dance’, TAPA 85,
148–58.
Letrouit, J. (1991) ‘À propos de la tradition manuscrite d’Athénée: une mise
au point’, Maia 43, 33–40.
LeVen, P. (2013) ‘The Colours of Sound: Poikilia in its Aesthetic Contexts’,
GRMS 1, 229–42.
LeVen, P. (2014) The Many-Headed Muse, Cambridge.
Levin, F. R. (1961) ‘The Hendecachord of Ion of Chios’, TAPA 92, 295–307.
LfgrE = B. Snell, et al. (eds) (1955–2010) Lexikon des frühgriechischen Epos,
Göttingen.
LIMC = L. Kahil et al. (eds) (1981–2009) Lexicon iconographicum mytholo-
giae classicae, Zurich.
Liverani, M. (1990) Prestige and Interest: International Relations in the Near
East ca. 1600–1100 B.C., Padua.
Lloyd-Jones, H. (1996). Sophocles, Fragments (Loeb). Cambridge, MA.
Lodato, S., Aspden, S., and Bernhart, W. (eds) (2002) Word and Music
Studies 4, Amsterdam.
London, J. (2002) ‘Cognitive Constraints on Metric Systems: Some Obser-
vations and Hypotheses’, Music Perception: An Interdisciplinary Journal
19.4, 329–50.
Long, A. A. (1995) ‘Cicero’s Plato and Aristotle’, in J. G. F. Powell (ed.), 37–61.
Lopes, R. (2009) ‘The Omnipresence of Philosophy in Plutarch’s Quaestiones
convivales’, in J. Ferreira, D. F. Leão, M. Tröster, and P. B. Dias (eds)
(2009), 415–24.
Loraux, N. (2000) Born of the Earth: Myth and Politics in Athens
(tr. S. Stewart), Ithaca, NY.
268 Bibliography
Lord, A. B. (2000) The Singer of Tales (2nd edn), Cambridge, MA.
Lord, C. (1982) Education and Culture in the Political Thought of Aristotle,
Ithaca, NY.
Lord, C. (1984) Aristotle, The Politics, Chicago.
Lynch, T. (2013) ‘Training the Soul in Excellence: Musical Theory and
Practice in Plato’s Dialogues, between Ethics and Aesthetics’, PhD Diss.,
St. Andrews.
Lynch, T. (2016) ‘Why are only the Dorian and Phrygian Harmoniai
Accepted in Plato’s Kallipolis? Lyre vs. Aulos’, in L. Bravi, L. Lomiento,
A. Meriani, and G. Pace (eds), 267–84.
Maas, M. and Snyder, J. (1989) Stringed Instruments of Ancient Greece, New
Haven and London.
Magie, D. (1967) Scriptores Historia Augustae (Loeb), Cambridge, MA.
Markantonatos, A. (ed.) (2012) The Brill Companion to Sophocles, Leiden.
Marshall, C. W. (2000) ‘Alcestis and the problem of prosatyric drama’, CJ 95,
229–38.
Martin, R. (1989) The Language of Heroes, Ithaca, NY.
Maslowski, T. (1978) ‘Cicero, Philodemus, Lucretius’, Eos 66, 215–26.
Mastronarde, D. (2002) Euripides, Medea, Cambridge.
Mastronarde, D. (2010) The Art of Euripides: Dramatic Technique and Social
Context, Cambridge.
Mathiesen, T. J. (1999) Apollo’s Lyre: Greek Music and Music Theory in
Antiquity and the Middle Ages, Lincoln, NE and London.
Maurey, Y., Seroussi, E., and Westenholz, J. (eds) (2014) Music in Antiquity,
Jerusalem and Berlin.
McClure, L. (ed.) (2017) Blackwell Companion to Euripides, Chichester.
Millis, B. W. and Olson, S. D. (2012) Inscriptional Records for the Dramatic
Festivals of Athens, Leiden.
Mirelman, S. (ed.) (2010) The Historiography of Music in Global Perspective,
Piscataway, NJ.
Mojsik, T. (2011) Between Tradition and Innovation: Genealogy, Names and
the Number of the Muses, Warsaw.
Morgan, K. A. (2015) Pindar and the Construction of Syracusan Monarchy in
the Fifth Century B.C., Oxford.
Morgan, T. J. (1998) Literate Education in the Greek and Roman Worlds,
Cambridge.
Morris, S. P. (1992) Daidalos and the Origins of Greek Art, Princeton, NJ.
Morrison, A. (2012) ‘Performance, Re–performance, and Pindar’s Audi-
ences’ in P. Agócs, C. Carey, and R. Rawles (eds), 111–33.
Mossman, J. (2011) Euripides: Medea, Oxford.
Mota, M. (ed.) (2010) Dramaturgias de cena, Brasilia.
Moutsopoulos, E. (1989) La Musique dans l’oeuvre de Platon (2nd edn),
Paris.
Mullen, W. (1982) Choreia: Pindar and Dance, Princeton, NJ.
Bibliography 269
Murnaghan, S. (2013) ‘The Choral Plot of Euripides’ Helen’. In R. Gagné and
M. Hopman (eds), 155–77.
Murray, P. and Wilson, P. (2004) Music and the Muses: the Culture of
Mousikē in the Classical Athenian City, Oxford.
Naerebout, F. (1997) Attractive Performances: Ancient Greek Dance; Three
Preliminary Studies, Amsterdam.
Nagy, G. (1990) Pindar’s Homer: The Lyric Possession of an Epic Past, Baltimore,
MD.
Nagy, G. (1996) Poetry as Performance: Homer and Beyond, Cambridge.
Nardelli, M. L. (ed.) (1983) ‘Due trattati filodemei Sulla Poetica’, in
F. Sbordone (ed.), Ricerche Sui Papiri Ercolanesi 4, Naples.
Nehamas, A. and Woodruff, P. (1995) Phaedrus. Indianapolis, IN.
Nettl, B. (1985) The Western Impact on World Music: Change, Adaptation,
and Survival, New York and London.
Nettl, B. (2005) The Study of Ethnomusicology (2nd edn), Champaign, IL.
Newman, W. L. (1902) The Politics of Aristotle, vol. 3, Oxford.
Newmyer, S. T. (2014) ‘Animals in Plutarch’, in M. Beck (ed.) A Companion
to Plutarch, 223–34, Chichester.
Nicholson, N. (2016) The Poetics of Victory in the Greek West, Oxford and
New York.
Nietzsche, F. W. (1967–) Kritische Gesamtausgabe: Werke, ed. G. Colli and
M. Montinari, 247–69, Berlin.
Nightingale, A. (1996) ‘Aristotle on the “Liberal” and “Illiberal” Arts’, Proceedings
of The Boston Area Colloquium in Ancient Philosophy 12, 29–58.
Nightingale, A. (2004) Spectacles of Truth in Classical Greek Philosophy,
Cambridge.
Nilsson, M. P. (1932) The Mycenaean Origin of Greek Mythology, Cambridge.
Nimis, S. A. (2007) ‘Autochthony, Misogyny and Harmony: Medea 824–45’,
Arethusa 40, 397–420.
Nooter, S. (2012) When Heroes Sing: Sophocles and the Shifting Soundscape of
Tragedy, Cambridge.
Norden, E. (1971 [1909–18]) Die antike Kunstprosa, Darmstadt.
Obbink, D. (ed.) (1995) Philodemus and Poetry: Poetic Theory and Practice in
Lucretius, Philodemus, and Horace, New York.
Oikonomopoulou, K. (2011) ‘Peripatetic Knowledge in the Table Talk’, in
F. Klotz and K. Oikonomopoulou (eds) (2011), 105–30.
Olson, S. D. (2006–12) Athenaeus: The Learned Banqueters (Loeb),
Cambridge, MA.
Osborne, R. (ed.) (2007) Debating the Athenian Cultural Revolution, Cambridge.
Padel, R. (1974) ‘ “Imagery of the Elsewhere:” Two Choral Odes of Euripides’,
CQ 24, 227–41.
Pagliara, A. (2000) ‘Musica e politica nella speculazione platonica: consider-
azioni intorno all’ethos del modo frigio (Resp. III 10, 399a–c)’, in A. Cassio,
D. Musti, and L. Rossi (eds), 157–216.
270 Bibliography
Pangle, T. L. (1980) The Laws of Plato, New York.
Panofsky, E. (1924) ‘Idea’: Ein Beitrag zur Begriffsgeschichte der älteren
Kunsttheorie, Leipzig.
Parker, L. (2007) ‘Dionysius’ Ear’, in P. Finglass, C. Collard, and
N. Richardson (eds), 297–305.
Pavese, C. O. (1997) ‘Sulla “Thebais” di Stesicoro’, Hermes 125.3, 259–68.
Pelosi, F. (2010) Plato on Music, Soul, and the Body (tr. S. Henderson),
Cambridge.
Pender, E. (2007) ‘Sappho and Anacreon in Plato’s Phaedrus’, Leeds Inter-
national Classical Studies 6, 1–57.
Peponi, A.-E. (2009) ‘Choreia and Aesthetics in the Homeric Hymn to Apollo:
The Performance of the Delian Maidens (Lines 156–64)’, ClAnt 28.1: 39–70.
Peponi, A.-E. (2012) Frontiers of Pleasure, Oxford.
Peponi, A.-E. (2013a) (ed.) Performance and Culture in Plato’s Laws,
Cambridge.
Peponi, A.-E. (2013b) ‘Choral Anti-Aesthetics’, in A.-E. Peponi (ed.),
212–39.
Peponi, A.-E. (2015) ‘Dance and Aesthetic Perception’, in P. Destrée and
P. Murray (eds), 204–17.
Pfister, F. (1909–12) Der Reliquienkult im Altertum, Giessen.
Phillips, T. (2013) ‘Epinician Variations: Music and Text in Pindar, Pythians
2 and 12’, CQ 63.1, 37–56.
Phillips, T. (2015) ‘Echo in Euripides’ Andromeda’, GRMS 3, 53–66.
Phillips, T. (2016) Pindar’s Library: Performance Poetry and Material Texts,
Oxford.
Picken, L. (1975) Folk Musical Instruments of Turkey, Oxford.
Pohl, K. (1968) ‘Die Lehre von den drei Wortfügungsarten: Untersuchungen
zu Dionysios von Halikarnaß, De compositione verborum’. Diss.,
Tübingen.
Pöehlmann, E. (1970) Denkmäler altgriechischer Musik, Nürnberg.
Pöehlmann, E. (1971) ‘Die ABC-Komödie des Kallias’, RhM 114, 230–40.
Pöehlmann, E. (1995) ‘Metrica ritmica nella poesia e nella musica greca
antica’, in B. Gentili and F. Perusino (eds), 3–15.
Pöehlmann, E. and Speliopoulou, I. (2007) ‘DAGM 27: Ὁ Ὕμνος στὸν Ἥλιο
τοῦ Μεσομήδη’, in Ἡ ἀρχαία ἑλληνικὴ μουσικὴ στὸ πλαίσιο τῆς ἀρχαίας
ἑλληνικῆς ποίησης, Kerkyra, 139–40.
Poliakoff, M. B. (1987) Combat Sports in the Ancient World, New Haven and
London.
Pomeroy, S. B. (1977) ‘Technikai kai mousikai: The Education of Women in
the Fourth Century and in the Hellenistic Period’, AJAH 2, 51–68.
Porter, J. I. (1989) ‘Philodemus on Material Difference’, Cron. Erc. 19, 149–78.
Porter, J. I. (1992) ‘Hermeneutic Lines and Circles: Aristarchus and Crates
on Homeric Exegesis’, in R. Lamberton and J. Keaney (eds), 67–114.
Bibliography 271
Porter, J. I. (1993) ‘The Seductions of Gorgias’, ClAnt 12.2, 267–99.
Porter, J. I. (1994) ‘Stoic Morals and Poetics in Philodemus’, Cron. Erc. 24,
63–88.
Porter, J. I. (1995a) ‘οἱ κριτικοί: A Reassessment’, in J. Abbenes, S. Slings, and
I. Sluiter (eds), 83–109.
Porter, J. I. (1995b) ‘Content and Form in Philodemus: The History of an
Evasion’, in D. Obbink (ed.), 97–147.
Porter, J. I. (2001) ‘Ideals and Ruins: Pausanias, Longinus, and the Second
Sophistic’, in S. Alcock, J. Cherry, and J. Elsner (eds), 63–92.
Porter, J. I. (ed.) (2005a) Classical Pasts, Princeton, NJ.
Porter, J. I. (2005b) ‘Feeling Classical: Classicism and Ancient Literary
Criticism’, in J. I. Porter (ed.), 301–52.
Porter, J. I. (2006) ‘Feeling Classical: Classicism and Ancient Literary Criti-
cism’, in J. I. Porter (ed.) Classical Pasts: the Classical Traditions of Greece
and Rome, Princeton, 301–52.
Porter, J. I. (2010) The Origins of Aesthetic Thought in Ancient Greece,
Cambridge.
Porter, J. I. (2016) The Sublime in Antiquity, Cambridge.
Powell, J. G. F. (ed.) (1995) Cicero the Philosopher, Oxford and New York.
Power, T. (2007) ‘Ion of Chios and the Politics of Polychordia’, in V. Jennings
and A. Katsaros (eds), 179–205.
Power, T. (2010) The Culture of Kitharodia, Cambridge, MA and London.
Power, T. (2012a) ‘The Power of Music: the Lyric Sophocles’, in
A. Markantonatos (ed.), 283–304.
Power, T. (2012b) ‘Aristoxenus and the “Neoclassicists” ’, in C. Huffman
(ed.), 129–54.
Power, T. (2013) ‘Kyklops kitharoidos: Dithyramb and Nomos in Play’, in
B. Kowalzig and P. Wilson (eds), 237–56.
Power, T. (forthcoming) ‘New Music in Sophocles’ Ichneutai’ in R. Andújar
et al. (eds) Paths of Song, Berlin.
Prauscello, L. (2006) Singing Alexandria: Music Between Practice and Textual
Transmission, Leiden and Boston.
Prauscello, L. (2013a) Performing Citizenship in Plato’s Laws, Cambridge.
Prauscello, L. (2013b) ‘Choral Persuasion in Plato’s Laws’, in R. Gagné and
M. G. Hopman (eds), 257–77.
Psaroudakes, S. (2010) ‘The Enunciation of “Metra” in Ancient Hellenic
Tragedy: Case in Point, Aischylos’ Agamemnon, ll. 40–6’, in M. Mota
(ed.), 59–72.
Pucci, P. (1977) Hesiod and the Language of Poetry, Baltimore, MD.
Pucci, P. (1979) ‘The Song of the Sirens’, Arethusa 12.2, 121–32.
Pucci, P. (1980) The Violence of Pity in Euripides’ Medea, Ithaca, NY.
Pucci, P. (1987) Odysseus Polutropos: Intertextual Readings in the Odyssey
and the Iliad, Ithaca, NY.
272 Bibliography
Raad, V. (2005) ‘Folklore and reminiscence in Claude Debussy’, in
P.-A. Bempéchat (ed.), Liber Amicorum Isabelle Cazeaux, Hillsdale, NY.
Rashid, S. A. (1984) Mesopotamien, Leipzig.
Reese, G. (1941) Music in the Middle Ages, London.
Revermann, M. and Wilson, P. (2008) (eds), Performance, Iconography,
Reception: Studies in Honour of Oliver Taplin, Oxford.
Reichl, K. (2000) The Oral Epic: Performance and Music, Berlin.
Richter, L. (1968) ‘Die neue Musik der griechischen Antike’, AfMW 25, 1–18,
134–47.
Richter, W. (1974) ‘συμφωνία: Zur Vor- und Frühgeschichte eines musiko-
logischen Begriffs’, in H. Hüschen and D.-R. Moser (eds), 264–90.
Rix, H. (1991) Etruskische Texte, vol. 2, Tübingen.
Rocconi, E. (2003) Review of Pöhlmann and West (2001), in CR 53.1, 69–71.
Rocconi, E. (2012) ‘The Aesthetic Value of Music in Platonic Thought’, in
I. Sluiter and R. Rosen (eds), 113–32.
Rodríguez-Noriega Guillén, L. (2000) ‘Are the Fifteen Books of the Deip-
nosophists an Excerpt?’, in D. Braund and J. Wilkins (eds), 244–55.
Roman, Z. (1973) ‘Connotative Irony in Mahler’s Todtenmarsch in ‘Callots
Manier’, Musical Quarterly 59, 207–22.
Rosen, C. (1998) The Classical Style, New York.
Rosen, R. M. (1999) ‘Comedy and Confusion in Callias’ Letter Tragedy’, CPh
94, 147–67.
Roskam, G. (2009) ‘Educating the young . . . over wine? Plutarch, Calvenus
Taurus and Favorinus as Convivial Teachers’, in J. Ferreira, D. F. Leão,
M. Tröster, and P. B. Dias (eds), 369–84.
Rossi, L. E. (1988) ‘La dottrina dell’ ethos musicale e il simposio’, in B. Gentili
and R. Pretagostini (eds), 238–45.
Ruffy, M. V. (2012) Les Vertus Thérapeutiques du Banquet, Paris.
Ruijgh, C. J. (1961) ‘Le Traitement des sonantes voyelles dans les dialectes
grecs et la position du mycénien’, Mnem. 14, 193–216.
Ruijgh, C. J. (2001) ‘Le Spectacle des lettres, comédie de Callias (Athénée
X 453c–455b), avec un excursus sur les rapports entre la mélodie du chant
et les contours mélodiques du langage parlé’, Mnem. 54, 257–335.
Russell, D. A. and Winterbottom, M. (1972) Ancient Literary Criticism, Oxford.
Russo, T. (ed.) (2012) Recognition and Modes of Knowledge, Edmonton, AB.
Salecl, R. (1998) (Per)versions of Love and Hate, London.
Sandbach, F. H. (1982) ‘Plutarch and Aristotle’, ICS 7, 207–32.
Santé, P. (2008) Gli scoli metrici a Pindaro, Pisa and Rome.
Saunders, T. J. (1970) Plato: The Laws. Harmondsworth.
Saunders, T. J. (1997) Plato, Laws, in J. Cooper (ed.), Plato: Complete Works,
Indianapolis, IN.
Sbordone, F. (ed.) (1976) Tractatus Tres in F. Sbordone (ed.), Ricerche sui
Papiri Ercolanesi, vol. 2, Naples.
Bibliography 273
Scher, S. (1997) ‘Melopoetics Revisited’ in W. Bernhart, S. Scher, and
W. Wolf (eds), 9–24.
Schoen-Nazzaro, M. B. (1978) ‘Plato and Aristotle on the Ends of Music’,
Laval théologique et philosophique 34, 261–73.
Schur, D. (2014) ‘The Silence of Homer’s Sirens’, Arethusa 47.1, 1–17.
Schütrumpf, E. (2005) Aristoteles Politik VII–VIII, Berlin.
Scott, G. (2005) ‘Twists and Turns: Modern Misconceptions of Peripatetic
Dance Theory’, Dance Research 23.2, 153–72.
Scott, W. C. (1984) Musical Design in Aeschylean Theater, London.
Scott, W. C. (1996) Musical Design in Sophoclean Theater, London.
Scruton, R. (1976) ‘Representation in Music’, Philosophy 51, 273–87.
Sfyroeras, P. (1994/5) ‘The Ironies of Salvation: The Aigeus Scene in Euripi-
des’ Medea’, CJ 90, 125–42.
Shaffer, A. (1981) ‘A New Musical Term in Ancient Mesopotamian Music’,
Iraq 43, 79–83.
Shaw, C. (2014) Satyric Play: The Evolution of Greek Comedy and Satyr
Drama, Oxford.
Shepherd, J. and Wicke, P. (1997) Music and Cultural Theory, London.
Silk, M. (2007) ‘Pindar’s Poetry as Poetry: A Literary Commentary on
Olympian 12’ in S. Hornblower and C. Morgan (eds), 177–97.
Slater, N. W. (2002) ‘Dancing the Alphabet: Performative Literacy on the
Attic Stage’, in I. Worthington and J. Foley (eds) 117–29.
Slater, N. W. (2005) ‘Nothing to Do with Satyrs?’: Alcestis and the Concept of
Prosatyric Drama’, in G. Harrison (ed.), Satyr Drama: Tragedy at Play,
83–101, Swansea.
Sluiter, I. and Rosen, R. (eds) (2012) Aesthetic Value in Classical Antiquity,
Leiden.
Smethurst, M. (2002) ‘Ninagawa’s Production of Euripides’ Medea’, AJP 123,
1–34.
Smith, B. R. (1999) The Acoustic World of Early Modern England: Attending
to the O-Factor, Chicago.
Smith, J. A. (2003) ‘Clearing Up Some Confusion in Callias’ Alphabet
Tragedy’, CP 98, 313–29.
Snyder, J. M. (1989) The Woman and the Lyre: Women Writers in Classical
Greece and Rome, Carbondale, IL.
Sobak, R. (2013) ‘Dance, Deixis, and the Performance of Kyrenean
Identity: A Thematic Commentary on Pindar’s Fifth Pythian’, HSCP
107, 99–153.
Solmsen, F. (1964) ‘Leisure and Play in Aristotle’s Ideal State’, RhM 107,
193–220.
Solomon, J. (1986) ‘The Seikilos Inscription: A Theoretical Analysis’, AJP
107.4, 455–79.
Spengel, L. (1828) συναγωγὴ τεχνῶν, sive Artium scriptores, Stuttgart.
274 Bibliography
Steiner, D. (2011) ‘Dancing with the Stars: Choreia in the Third Stasimon of
Euripides’ Helen’, CP 106, 299–323.
Steiner, D. (2015) ‘The Poetics of Sound: Callimachus’ Rereading of Pindar
Fragment 70B S.-M.’, CP 110, 99–123.
Steiner, G. and Fagles, R. (eds) (1962) Homer: A Collection of Critical Essays,
Englewood Cliffs, NJ.
Stewart, S. (2002) Poetry and the Fate of the Senses, Chicago.
Stockert, W. (1969) Klangfiguren und Wortresponsionen bei Pindar, Vienna.
Sutton, D. G. (1983) ‘The Date of Prometheus Bound’, GRBS 24, 289–94.
Swift, L. (2009) The Hidden Chorus, Oxford.
Teodorsson, S.-T. (1996) A Commentary on Plutarch’s Table Talks, vol. 3,
Göteborg.
Thaut, M. (2005) Rhythm, Music, and the Brain: Scientific Foundations and
Clinical Applications, London.
Thorp, J. (1991) ‘Aristoxenus and the Ethnoethical Modes’, in R. Wallace
and B. MacLachlan (eds), 54–65.
Too, Y. L. (1998) The Idea of Ancient Literary Criticism, Oxford.
Torrance, I. (2013) Metapoetry in Euripides, Oxford.
Tsitsiridis, S. (2013) Beiträge zu den Fragmenten des Klearchos von Soloi, Berlin.
Vazaki, A. (2003) Mousike Gyne: die musisch-literarische Erziehung und
Bildung von Frauen im Athen der klassischen Zeit, Möhnesee.
Vermeule, E. (1979) Aspects of Death in Early Greek Art and Poetry, Berkeley.
Vernant, J-P. (1974) Mythe et société en Grèce ancienne, Paris.
Vox, O. (2003) ‘Euripide, Medea 190–203’, in F. Benedetti and S. Grandolini
(eds), 829–35.
Walden, J. (ed.) (2013) Representation in Western Music, Cambridge.
Wallace, R. and MacLachlan, B. (eds) (1991) Harmonia Mundi, Rome.
Wallrup, E. (2015) Being Musically Attuned: The Act of Listening to Music,
Farnham.
Węcowski, M. (2014) The Rise of the Greek Aristocratic Banquet, Oxford.
Wehrli, F. (1948) Die Schule des Aristoteles: III Klearchos, Basel.
Weiss, N. (2008) ‘A Psychoanalytical Reading of Euripides’ Ion: Repetition,
Development and Identity’, BICS 51, 39–50.
Weiss, N. (2012) ‘Recognition and Identity in Euripides’ Ion’, in T. Russo
(ed.), 33–50.
Weiss, N. (2018) The Music of Tragedy: Performance and Imagination in
Euripidean Theatre, Berkeley.
Weiss, N. (forthcoming) ‘Performing the hymenaios in Euripides’ Iphigenia
in Aulis’, in R. Andújar et al. (eds).
Welcker, F. G. (1832) ‘Das A.B.C. Buch des Kallias in Form einer Tragödie’,
RhM 1, 137–57.
West, M. L. (1966) Hesiod: Theogony, Oxford.
West, M. L. (1981) ‘The Singing of Homer and the Modes of Early Greek
Music’, JHS 101, 113–29.
Bibliography 275
West, M. L. (1982) Greek Metre, Oxford.
West, M. L. (1992a) ‘Analecta Musica’, ZPE 92, 1–54.
West, M. L. (1992b) Ancient Greek Music, Oxford.
West, M. L. (1993–4) ‘The Babylonian Musical Notation and the Hurrian
Melodic Texts’, Music and Letters 75, 161–79.
West, M. L. (1997) The East Face of Helicon, Oxford.
West, M. L. (2000) ‘Music Therapy in Antiquity’, in P. Horden (ed.), 51–68.
West, M. L. (2011) Collected Papers, vol. 2, Oxford.
Westenholz, J. G. (ed.) (2007) Sounds of Ancient Music, Jerusalem.
Westphal, R. (1861) Die Fragmente und die Lehrsätze der griechischen
Rhythmiker, Leipzig.
Widdess, R. (1995) The Ragas of Early Indian Music, Oxford.
Wilamowitz-Möllendorff, U. v. (1903) Timotheus: Die Perser, Leipzig.
Wilson, P. (1999) ‘The Aulos in Athens’, in S. Goldhill and R. Osborne (eds),
58–95.
Wilson, P. (1999–2000) ‘Euripides’ Tragic Muse’ in M. Cropp, K. Lee, and
D. Sansone (eds), 427–49.
Wilson, P. (2000) The Athenian Institution of the Khoregia, Cambridge.
Wilson, P. (2003) ‘The Sound of Cultural Conflict: Kritias and the Culture of
Mousikê in Athens’, in C. Dougherty and L. Kurke (eds), 181–206.
Wilson, P. (2005) ‘Music’, in J. Gregory (ed.), A Companion to Greek
Tragedy, Oxford, 183–93.
Wilson, P. (2008) ‘Music’, in J. Gregory (ed.), 183–94.
Wilson, P. (2009) ‘Thamyris the Thracian: The Archetypal Wandering
Poet?’, in R. Hunter and I. Rutherford (eds), 46–79.
Winnington-Ingram, R. P. (1936) Mode in Ancient Greek Music, Cambridge.
Winnington-Ingram, R. P. (1955) ‘Fragments of Unknown Greek Tragic
Texts with Musical Notation: II, The Music’, SO 31, 29–87.
Winnington-Ingram, R. P. (1963) Aristides Quintilianus, De Musica, Leipzig.
Wolf, W. (2002) ‘Intermediality Revisited: Reflections on Word and Music
Relations in the Context of a General Typology of Intermediality’, in
S. Lodato, S. Aspden, and W. Bernhart (eds), 13–34.
Worthington, I. and Foley, J. (eds) (2002) Epea and Grammata, Leiden.
Wulstan, D. (1968) ‘The Tuning of the Babylonian Harp’, Iraq 30, 215–28.
Yatromanolakis, D. (2007) Sappho in the Making: The Early Reception,
Washington, DC.
Yatromanolakis, D. (ed.) (2011) Music and Cultural Politics in Greek and
Chinese Societies, vol. 1, Boston, MA and London.
Zaminer, F. (1984) ‘Hypate, Mese und Nete in frühgriechischen Denken: Ein
Altes musikterminologisches Problem in neuem Licht’, AfMW 41, 1–26.
Ziegler, N. (2007) Les Musiciens et la musique d’après les archives royals de
Mari, Paris.
Žižek, S. (1997) The Abyss of Freedom, Ann Arbor.
Index
Aelian 49 Dionysus 153

Aeschylus 105, 113 n. 37, 221 theatre of 140, 142
aesthetics 48, 167–8, 204, 212,
216, 230 emotion 10–11, 61–2, 127–8, 192–3,
affectivity 75–9, 81–2, 167 196–7, 199–200, 207, 225, 234
Apollo 28, 30, 44, 46, 89, 91, 141 n. 12, ethos 48
156–9 Etna (volcano) 89–97
appearances 219, 221, 228 Euripides 58–64, Ch. 4 passim, Ch. 6
Aristides Quintilianus 77–9, 123 n. 11 passim, 220, 238 n. 5
Aristotle 25–6, 41, 48, 50 n. 14,
80 n. 21, 148, 177, Ch. 8 passim, Gorgias 204 n. 1, 208, 210, 217, 237
217, 241–7 Griffith, M. 159
Aristoxenus 20, 28, 38, 48, 105 n. 18, Gumbrecht, H. U. 168
141, 217–18, 231 n. 95, 241, 243
arsis and thesis 63–4 Hagel, S. 40–1, 52–3
Athena 147, 156, 200–1 harmonia (ἁρμονία) 31
Athenaeus 100–9 heptachord Ch. 1 passim
atmosphere see also ‘Stimmung’ 5 n. 19, Hermes 30, 43, 44 n. 117, 45, 140,
11, 128 n. 29, Ch. 7 passim 143, 152
aulos 5, 10, 28, 49, Ch. 6 passim, 200, Homer 2, 34, 51–7, 84–6, 165–6,
239, 245 n. 19, 250–1 189–91, 204, 208–11
Hurrian hymns 24, 32
Barker, A. 160
Burkert, W. 28–9 Ihde, D. 144
intelligence (φρόνησις) 200–1
Callias (‘Alphabet Drama’) 100–9, 237 intermediality 3–5, 249–55
chorus 7, Ch. 6 passim, 235–6 intervals
choral dance 79, 95 n. 58 melodic 20–4, 38 n. 90, 117, 128 n. 27
choral singing 58–64, 102–4, 112–20 rhythmic 74, 204, 213, 218–22
rehearsal of 81 n. 22
star chorus 130 judgement (κρίνειν) 193–5
Cicero 204–20, 229–30
Collard, C. 160 Kinyras 18, 30
Cook, N. 87–8 kithara 32 n. 61, 140–1, 149, 152,
Crates of Malos 222–6, 228–9 156–9, 199
Creusa 150–1, 162 Kovacs, D. 160
Csapo, E. 100, 144, 147, 152 Kroll, W. 211
Debussy, C. 139–40 LeVen, P. A. 151–2

Demosthenes 207, 211, 216, 229 Longinus 49, Ch. 9 passim
diatonic tuning lyre Ch. 1 passim, 54, 58, 113, 115, 123,
technical description of 19–27 140–2, 152, 157, 161, 185, 197, 226–7
Mesopotamian origins of 27–35
Dionysius of Halicarnassus 48, 74–5, 78, mediality 3–5
79 n. 18, 90, 214–16, 218, 220–5 see also ‘intermediality’
278 Index
melody (μέλος) pitch accent
less central than harmonia 50 in Homeric hexameters 51–3
emotive effects 58–64, Ch. 5 passim, and melody 37–9, Ch. 2 passim, 105,
esp. 127–9, 173 118–19
imitative use of 130 plainsong, Gregorian 65, 72
sensuous effects of 174–8 Plato Ch. 7 passim
and responsion 112–18 and the sublime 229
Mesomedes 66, Ch. 5 passim Aristotle’s responses to Ch. 8 passim
metre influence on Cicero 206–10
apocrota 125–6 influence on Plutarch Ch. 10 passim,
dochmiacs 63 esp. 236–43
and stanzaic responsion 89–97, on ἁρμονία 228–9
104–5, 112 on dance 80–1
and rhythm 62–3, 68–72, 76–80 on lyre tuning 25, 28, 44
Plato’s views on 169–74 on the aulos 148–9, 157
see also ‘rhythm’ on the ‘New Music’ 111 n. 33, 147–8
mimesis on rhythm 170–2, 175–6
and representation 130, 174–8 Plutarch 44–5, Ch. 10 passim
and rhythm 87–9, 95–6 ποικιλία 151–2, 187 n. 6
of emotions 192 Protagoras 236–7
in dance 80
instrumental 140, 147, 159, 161 responsion see ‘metre’ and ‘melody’
Morwood, J. 160 rhythm
Muses 44–5, 55, 89, 91, 95–7, 114–15, affective force of 76–9, 81–4, 170–2
116, 142, 147 Aristotle’s views on 192, 194, 196
Euripides’ use of 58–9, 64
‘New Music’ 28, 32–3, 58–9, 64, 99–100, in ancient critics 49–50, 74–81, 213,
142, 145, 147, 152, 159, 164, 175, 215, 217–20
177–8, 187, 199–200 in Homer 55
Mesomedes’ use of 126–9
Orpheus 113 Pindar’s use of Ch. 3 passim, esp.
παίδεια Ch. 8 passim 81–7, 91–8
Plato’s views on 170–2, 175–6
perception Seikilos’ use of 68–71
and melody 105, 115 n. 44 rhythmical enactment 75–6, 81–7, 93–5
and prosodic texture 167–8, 174, riddles 108
180–1
and rhythm 83, 88–9 Sappho 37, 39, 49, 90, 114, 165–7,
performance 179–80, 207–8
and epicentric tonality 35–40 Seikilos 41–2, 64–72, 121, 130
and mimesis 87, 96–7, 130, Ch. 6 singing
passim, esp. 140–2, 148, 156–7 and the voice 216
Aristotle’s views on 185, 193–5 choral 58–64, 79 n. 18, 114–18, 151,
ethical implications of Ch. 10 passim, 155, 156
esp. 236–42 in education 185, 186 n. 5
Plato’s views on 169–78 of hexameters 28–9, 51–7
Philodemus 76 n. 9, 94 n. 57, of lyric 38–9
211–13, 230 to onself (μινυρίζειν) 173
Phrynicus of Athens 48 Sisyphus 74
Phrynis 33 Sophocles 100, 101–4, 140, 143, 152
Pindar 46, Ch. 3 passim, 105, 140, 147, stanzaic interaction 89–97, 112–18
221, 231 n. 97 Stimmung 5, 167–8, 172–4, 178
Index 279
sublimity 49, 96, Ch. 9 passim voice Ch. 9 passim
σχολή Ch. 8 passim and singing 5–6, 123, 128 n. 29, 173–4
syrinx 10, 100 n. 2, Ch. 6 passim, of instruments 150
204 n. 2
Webern, A. 229
Terpander 28–30, 33–4, 44 West, M. L. 28–9, 34, 35, 51–2
Timotheus 32 n. 61, 33 Westphal, R. 216
Tynnichus of Calchis 48
Typhos 93–7 Zeus 43, 81, 84, 89, 194

Music, Text, and Culture in Ancient Greece

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Music, Text, and Culture in Ancient Greece

Uploaded by

Copyright:

Available Formats

OUP CORRECTED PROOF – FINAL, 30/1/2018, SPi

MUSIC, TEXT, AND CULTURE

Music, Text, and

This volume originates in a workshop that took place in June 2013,

Introduction: Music, Text, and Culture 1

P ART II. THE ORY, R ECEP TION, CONTEXTS

CBS Catalogue of the Babylonian Section, University Museum,

Andrew Barker is Emeritus Professor of Classics at the University of

xiv List of Contributors

This volume addresses two issues central to the study of ancient

Introduction: Music, Text, and Culture 3

Introduction: Music, Text, and Culture 5

Introduction: Music, Text, and Culture 7

Introduction: Music, Text, and Culture 9

Introduction: Music, Text, and Culture 11

Introduction: Music, Text, and Culture 13

Epicentric Tonality and the Greek

In 1996, as a doctoral candidate at University College London, I set

Epicentric Tonality and the Greek Lyric Tradition 19

EPICENTRIC TONALITY BETWEEN GREECE

The discovery of a very early diatonic tuning/tonal system in Mesopo-

Franklin (2015); Chapter 5, Franklin (2003); Chapter 7, Franklin (2002d); Franklin

Epicentric Tonality and the Greek Lyric Tradition 21

(ἀρχή) of ‘continuity’ or ‘cohesion’ (συνέχεια) required that a properly

Epicentric Tonality and the Greek Lyric Tradition 23

Figure 1.2. Graphic representation of the Mesopotamian tuning cycle UET

2. A sequence of seven diatonic tunings, attested by the ‘Retuning

Epicentric Tonality and the Greek Lyric Tradition 25

Figure 1.3. Epicentric arrangement of traditional Greek heptachord, according

4. that μέση was often repeated in ‘worthy melodies’ (τὰ χρηστὰ

Epicentric Tonality and the Greek Lyric Tradition 27

Anyone who accepts some connection between the Mesopotamian

generally, including socio-political and literary dimensions, Richter (1968); Wilson

Epicentric Tonality and the Greek Lyric Tradition 29

Epicentric Tonality and the Greek Lyric Tradition 31

Epicentric Tonality and the Greek Lyric Tradition 33

Epicentric Tonality and the Greek Lyric Tradition 35

MUSIC, TEXT, AND EPICENTRIC TONALITY

This section offers preliminary observations on a pair of complemen-

Epicentric Tonality and the Greek Lyric Tradition 37

Epicentric Tonality and the Greek Lyric Tradition 39

Figure 1.4. ‘Recomposition’ of Sappho 1, illustrating epicentric tonality and

recomposition of Sappho 1, adhering to epicentric tonality, accent

Epicentric Tonality and the Greek Lyric Tradition 41

linear presentation of octaves species or the Perfect System in the musicographers.

Epicentric Tonality and the Greek Lyric Tradition 43

Figure 1.5. Didymus’ chromatic γένος, expressed in matrix showing ratios

From around the same time as Didymus come two formulations of

Epicentric Tonality and the Greek Lyric Tradition 45

double-entendre, ‘Who is this Muse who causes inescapable cares’—inverting the

The Musical Setting of Ancient

Scholarly examination of the few surviving ancient Greek musical

The Musical Setting of Ancient Greek Texts 49

The Musical Setting of Ancient Greek Texts 51

MUSICAL EFFECTS IN HOMER

Homer’s exploitation of the rhythmical features of the hexameter

The Musical Setting of Ancient Greek Texts 53

The Musical Setting of Ancient Greek Texts 55

πολλῶν δ᾽ ἀνθρώπων ἴδεν ἄστεα καὶ νόον ἔγνω,

The Musical Setting of Ancient Greek Texts 57

The centuries that followed Homer saw the introduction to Greece of

The Musical Setting of Ancient Greek Texts 59

The Musical Setting of Ancient Greek Texts 61